go
/
DateKeys
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '4f7885495b'
${ noResults }
DateKeys/testdata/vectors/mutations.json

4943 lines
971 KiB
Raw Normal View History Unescape

Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
{
v0.16: SpecVersion 0.16, the annex of the draft, and the tests in 76 SpecVersion is 0.16, and so is the spec field of every file of testdata, the frozen security_cms.json and locator.json included. annex/recovery.md is §79 of the draft v0.16, with the key of words in 79.7. The draft lists the two new fixtures in 67, says in 79 what recovery_check.sh opens, and names in 76 the tests that now exist. The CHANGELOG has its section. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 hours ago
"spec": "0.16",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"description": "Mutation corpus of spec §64 and further cases of capsule.TestMutationCorpus, generated by the reference implementation: each case is a .dkc and what the reader is given, with the normative error and the step of spec §63 at which capsule.Open fails. See testdata/README.md.",
"cases": [
{
"name": "PUBLIC_HEADER_A + SEALED_CONTROL_B",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_HEADER_BINDING",
"step": 15
},
{
"name": "SEALED_CONTROL_A + PAYLOAD_AGE_B",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "DateKey A + release of round B",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_ROUND_MISMATCH",
"step": 10
},
{
"name": "chain hash changed",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[173, 64, "61626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_PROFILE_MISMATCH",
"step": 8
},
{
"name": "version changed",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 3: the reader Open reads format 3 (spec 29.2 to 29.7, 63 steps 17 and 18): the PRELUDE accepts VERSION 3, and the files go to a Sink. - Sink: Begin with the validated head, Create for each file in the order of the head, and Commit only after every check of step 17; after any failure that follows a successful Begin, Abort, once. A format 3 capsule without a Sink fails right after step 2 with ErrSinkRequired, a caller error with no code, no failed step and no request; a capsule of format 1 or 2 without dst fails there too. - Step 17 in its substeps: the frame and the area, security and its verdicts, which never fail, the head, the files filling CONTENT, the SHA-256 of each file and the padding. A failure of age or a plaintext whose length is not P prevails; otherwise the first substep that fails decides, and a code other than ERR_INTEGRITY is reported only after reading PAYLOAD_AGE to its end. - The reads of BODY grow with the bytes received, never with AREA_LEN, HEAD_LEN or a declared size (spec 57); a test measures it. - A failure of the Sink is the caller's own error with ERR_INTEGRITY, as one of dst is in formats 1 and 2. - Opened gains Head, Verdicts, AreaLen and UnusableHeadExtensions. - Test data: "version changed" sets VERSION 4, and the format 2 list gains "format 2 time_only relabeled format 3", which fails at step 14, as section 64 of spec v0.10 lists: 126 cases, 89 of the spec. The randomly built capsules keep their recorded bytes. - testkit: Build writes format 3 and can edit the padded plaintext; Head3, Body3, DiscardSink and MemorySink build and open BODY. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[4, 1, "04"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 2
},
{
"name": "flags != 0",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[5, 1, "80"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_FLAGS",
"step": 2
},
{
"name": "reserved != 0",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[7, 1, "01"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_FLAGS",
"step": 2
},
{
"name": "payload truncated",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[78798, 1, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "payload age modified",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[78798, 1, "96"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "control modified",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[582, 1, "76"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "non-canonical dk1_ JSON",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 1, "7c"],
[51, 1, "56"],
[74, 61, "4169626d56306432397961794936496d5268644756725a586c7a4f6e463161574e72626d56304f6e597849697767496e4a766457356b496a6f784d4441"],
[135, 0, "776651"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_DATEKEY_NON_CANONICAL",
"step": 4
},
{
"name": "unknown profile",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 1, "76"],
[51, 1, "50"],
[100, 32, "5a585a74626d56304f6e597849697769636d3931626d51694f6a45774d444239"],
[132, 3, ""]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNKNOWN_PROFILE",
"step": 4
},
{
"name": "release of another round",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "access_policy=time_only with time_and_key structure",
"spec": true,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": [
[136, 1, "00"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b6579010102503955e944a3c60cfa1fd6485e9693c77d0350448e134a13457c319cab7fceaf7ffa1f04667832353531390558203d5e4d51f21236ecff40d89c0adcf9aa325a24b91b46fd9d1afa482f54b2d2c606a10058202e97878078bae6358037a9c264f379a3cbe839f767d69836b0343f35657b2972",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "access_policy=time_and_key with time_only structure",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[136, 1, "01"]
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "extra stanza in OUTER_TIME_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 5
},
{
"name": "extra stanza in PAYLOAD_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 6
},
{
"name": "non-X25519 stanza in INNER_ACCESS_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "tlock stanza round differs from DateKey.round",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_ROUND_MISMATCH",
"step": 8
},
{
"name": "tlock stanza chain hash differs from the pinned profile",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "444b43310100000000000079000001bea5006a646174656b6579636170010102505d6090f141146d36ba274b96b44eb941037853646b315f65794a325a584a7a61573975496a6f784c434a755a58523362334a72496a6f695a4746305a57746c65584d3663585670593274755a585136646a45694c434a79623356755a4349364d5441774d483004006167652d656e6372797074696f6e2e6f72672f76310a2d3e20746c6f636b203130303020646264353036643665663736653566333836663431633635316463623830386335626362643735343731636334656166613366346466376164346534633439330a6a776d34753269326c797a654b764d754268626954366f74642b5238635378614743442f6a3675534b662f6a2b2f52624c523872535a2b4567624955697544580a4369796c4264467865766c504b415a674c6b6e6547482f3930373248336d73443458326156345463464668474e50584a674e51474e782b75554d3544703775640a7032614c67752f387939636869474a78627148653533514f326c61496a33336f7770464543454349615a770a2d2d2d205464647a684338714142763251315a594c2b2b4d68766c453238424c5550393666654a446e6f4d72752b6f0abfc4d9ad388fb2d3699c69d385a97705e3e1a9dcea072e4d76a211e420ca3545c24cfdb5439cfc8e68e20db7034c7087b9f417cccbb6eed5d94482b75c3cf1d92d08b5eb0e36eb91fa8d02cff034c9c155cf47537395adb30821fa0784f0a434a114120b79ba0b42ff55766adfc9c0941aabee9775be3fd50655746167652d656e6372797074696f6e2e6f72672f76310a2d3e20583235353139204f756c766c764a4c706f562b3569334768383434685351354b45754d647473396f4f5753575153433253630a33665747774447696866467a2f7671767a77364d667a685254426875614e7151674d3548526a72696b376b0a2d2d2d2073695154577639354b635237576e48414f38547858746b7a45556272387458426c71614d514371347747630a2a9da42ad38e7ea56b45632f351cbf9cd9252ca816517b458154b4ce7c774a6e3df70375e73b07e26d95f783b9a45e020d"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_PROFILE_MISMATCH",
"step": 8
},
{
"name": "extension data of a type other than bstr",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "9f000001bea6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c0101026c7075626c6963206c6162656c"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "empty extension data (h'')",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "93000001bea6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c01010240"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "65 extensions in one array",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[10, 7, "0590000001bea6"],
[137, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
{
"name": "tlock stanza U re-encoded with c0 + p",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[303, 64, "4b3974334342577331793673785069396157564b384579504b612b796b6762676b4a46782f5a763371667465747a654b4f7a3852534856786467582f6e37506e"],
[416, 43, "675a464c773845307137726b6e384b675355414e76345a316679413258534e446c584653584b617456716f"]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "tlock stanza U is the point at infinity",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[238, 129, "774141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141410a41414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141"],
[416, 43, "472f6d426f3779325364374b6f4a567364384734547a357557346f6a77384e6467453772384b717a514534"]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "tlock stanza U with the infinity flag and a payload",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[238, 1, "77"],
[416, 42, "336a65787a7a2f586e7359324459467751754c706d6b6f2b6d465373366b786c6c775273487a6e51346e"]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "tlock stanza body of 127 bytes",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "bd"],
[409, 49, "410a2d2d2d206d6c44362b7a79424a6b524852363657776857654f5362584a5030394c6373734f57796c4962412f6c5649"],
[458, 1, ""]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "tlock stanza body of 129 bytes",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "bf"],
[411, 48, "410a2d2d2d20325a524d54626a4b62363170795949635756664f79336d75474f584f777056366b4d4b5a4b553434484a"],
[459, 0, "30"]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "release signature re-encoded with x + p",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1004,
"signature": "be076aa25a40df5b4d22f9f7bcedaff30dcac20d94556107b8e652c7b54b2a440ce796f9fa53ad28023c84b40a580f55"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "release signature is the point at infinity",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "c00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "release signature with the infinity flag and a payload",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "d44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "release signature negated",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "944679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "negated release signature and U re-encoded with c0 + p",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[303, 64, "4b3974334342577331793673785069396157564b384579504b612b796b6762676b4a46782f5a763371667465747a654b4f7a3852534856786467582f6e37506e"],
[416, 43, "675a464c773845307137726b6e384b675355414e76345a316679413258534e446c584653584b617456716f"]
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "944679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Spec v0.8.2 amendment: canonical point encoding; no library error text Amendment of the unreleased v0.8.2, recorded in §76 with its case: the second implementation's phase-2 research found that tlock-js over @noble/curves 1.9.7 accepts U re-encoded as c0 + p and a signature x + p and returns the same file key, while the reference rejects both (noble 1.9.7 differed from kilic on 5,615 of 41,686 encodings), and the spec did not say which encodings are valid. - §12.2 defines the canonical encoding of a BLS12-381 point (drand's compressed ZCash form) and requires decoders to reject every other byte string; §12.1 applies it to public_key. - §63 step 10 applies it to the release signature (ERR_RELEASE_INVALID) and step 11 defines the tlock stanza body U || V || W (96 + 16 + 16 bytes for Quicknet) with a canonical, non-infinity U (ERR_INTEGRITY). - §64 gains ten mutations, exported to mutations.json (65 cases). The signature x + p case uses published Quicknet round 1004, the first after 1000 whose x allows x + p < 2^381. The reference already gave every stated code and step. Errors no longer copy text from tlock, kyber, age, drand or kyber-bls12381. kyber's IBE error carried the candidate plaintext and r, and with one bit of W flipped the message disclosed the real tlock file key with that bit flipped. Every such place now uses a fixed reason with its normative sentinel; TestTlockFailureDiagnosticsCarryNoSecrets fails with the old wrapping. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
{
"name": "magic",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[0, 1, "58"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_MAGIC",
"step": 1
},
{
"name": "a .dkk offered as a .dkc",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[2, 1, "4b"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_MAGIC",
"step": 1
},
{
"name": "empty file",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[0, 78799, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_MAGIC",
"step": 1
},
{
"name": "truncated prelude",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[10, 78789, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 1
},
{
"name": "PUBLIC_HEADER_LEN above the limit",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[9, 3, "100001"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 2
},
{
"name": "SEALED_CONTROL_LEN above the limit",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[12, 4, "04000001"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 2
},
{
"name": "truncated inside SEALED_CONTROL",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[147, 78652, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 5
},
{
"name": "header schema version changed",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[30, 1, "02"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 4
},
{
"name": "unknown key in PUBLIC_HEADER",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "7b000001bea6"],
[136, 0, "0007"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "undefined access_policy",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[136, 1, "02"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "access_policy 256 with a consistent header_binding",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "access_policy 257 with a consistent header_binding",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "unknown critical PUBLIC_HEADER extension",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_EXTENSION_CRITICAL_UNKNOWN",
"step": 4
},
{
"name": "unknown critical CONTROL_CBOR extension",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_EXTENSION_CRITICAL_UNKNOWN",
"step": 14
},
{
"name": "known critical PUBLIC_HEADER extension with invalid data",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"extensions": [
{
"id": "org.example.must-understand",
"version": 1,
"valid_data": "6f6b"
}
],
"network": false,
"frozen": true,
"error": "ERR_EXTENSION_DATA_INVALID",
"step": 4
},
{
"name": "known critical CONTROL_CBOR extension with invalid data",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"extensions": [
{
"id": "org.example.must-understand",
"version": 1,
"valid_data": "6f6b"
}
],
"network": true,
"frozen": true,
"error": "ERR_EXTENSION_DATA_INVALID",
"step": 14
},
{
"name": "known critical .dkk extension with invalid data",
"spec": false,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": []
},
"dkk": "444b4b3101000000000000b3a80073646174656b6579732d6163636573732d6b6579010102503955e944a3c60cfa1fd6485e9693c77d0350448e134a13457c319cab7fceaf7ffa1f04667832353531390558203d5e4d51f21236ecff40d89c0adcf9aa325a24b91b46fd9d1afa482f54b2d2c606a10058202e97878078bae6358037a9c264f379a3cbe839f767d69836b0343f35657b29720781a300781b6f72672e6578616d706c652e6d7573742d756e6465727374616e64010102426b6f",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"extensions": [
{
"id": "org.example.must-understand",
"version": 1,
"valid_data": "6f6b"
}
],
"network": false,
"frozen": false,
"error": "ERR_EXTENSION_DATA_INVALID",
"step": 9
},
{
"name": "extension_version above 2^32-1",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "99000001bea6"],
[136, 0, "000681a200716f72672e6578616d706c652e6c6162656c011b00000001000000"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "null extension data",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "93000001bea6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c010102f6"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "time_and_key without credentials",
"spec": false,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_REQUIRED",
"step": 9
},
{
"name": ".dkk of another capsule",
"spec": false,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": []
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250b89292aedf6d05d584cec9a871ce87350350c75dfc8e9c576d1369910664df93693a046678323535313905582042d6d897097fd5af2772e058db17920afe1390e2856139bc2f800294564dd7ac06a100582069ac110380f5d768b5b6afaa157a50ed17d8ceccfbd4604ffa5b6da38539b635",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
},
{
"name": "capsule_digest of the .dkk does not match",
"spec": false,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": [
[1028, 1, "c1"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b6579010102503955e944a3c60cfa1fd6485e9693c77d0350448e134a13457c319cab7fceaf7ffa1f04667832353531390558203d5e4d51f21236ecff40d89c0adcf9aa325a24b91b46fd9d1afa482f54b2d2c606a10058202e97878078bae6358037a9c264f379a3cbe839f767d69836b0343f35657b2972",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
},
{
"name": "identity that is not a recipient",
"spec": false,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": []
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 13
},
{
"name": "round not reached yet",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:23.999999999Z",
"registry": "default",
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"network": true,
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"frozen": false,
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"error": "ok",
"step": 0
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
},
{
"name": "release source unavailable",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": null,
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_UNAVAILABLE",
"step": 9
},
{
"name": "trailing data after PAYLOAD_AGE",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[78799, 0, "00"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "payload stanza body modified",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[692, 1, "41"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "tlock round edited by a third party",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": [
[171, 1, "31"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ROUND_MISMATCH",
"step": 8
},
{
"name": "empty registry",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2023-08-23T15:59:24Z",
"registry": "empty",
"network": false,
"frozen": false,
"error": "ERR_UNKNOWN_PROFILE",
"step": 4
},
{
"name": "time_only declared, time_and_key built by the creator",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "time_and_key declared, time_only built by the creator",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "two INNER_ACCESS_AGE stanzas for one recipient",
"spec": false,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 13
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
},
{
"name": "format 2: PUBLIC_HEADER_A + SEALED_CONTROL_B",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_HEADER_BINDING",
"step": 15
},
{
"name": "format 2: SEALED_CONTROL_A + PAYLOAD_AGE_B",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 2: DateKey A + release of round B",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": []
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_ROUND_MISMATCH",
"step": 10
},
{
"name": "format 2: chain hash changed",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[173, 64, "61626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_PROFILE_MISMATCH",
"step": 8
},
{
"name": "format 2: version changed",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 3: the reader Open reads format 3 (spec 29.2 to 29.7, 63 steps 17 and 18): the PRELUDE accepts VERSION 3, and the files go to a Sink. - Sink: Begin with the validated head, Create for each file in the order of the head, and Commit only after every check of step 17; after any failure that follows a successful Begin, Abort, once. A format 3 capsule without a Sink fails right after step 2 with ErrSinkRequired, a caller error with no code, no failed step and no request; a capsule of format 1 or 2 without dst fails there too. - Step 17 in its substeps: the frame and the area, security and its verdicts, which never fail, the head, the files filling CONTENT, the SHA-256 of each file and the padding. A failure of age or a plaintext whose length is not P prevails; otherwise the first substep that fails decides, and a code other than ERR_INTEGRITY is reported only after reading PAYLOAD_AGE to its end. - The reads of BODY grow with the bytes received, never with AREA_LEN, HEAD_LEN or a declared size (spec 57); a test measures it. - A failure of the Sink is the caller's own error with ERR_INTEGRITY, as one of dst is in formats 1 and 2. - Opened gains Head, Verdicts, AreaLen and UnusableHeadExtensions. - Test data: "version changed" sets VERSION 4, and the format 2 list gains "format 2 time_only relabeled format 3", which fails at step 14, as section 64 of spec v0.10 lists: 126 cases, 89 of the spec. The randomly built capsules keep their recorded bytes. - testkit: Build writes format 3 and can edit the padded plaintext; Head3, Body3, DiscardSink and MemorySink build and open BODY. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[4, 1, "04"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 2
},
{
"name": "format 2: flags != 0",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[5, 1, "80"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_FLAGS",
"step": 2
},
{
"name": "format 2: reserved != 0",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[7, 1, "01"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_FLAGS",
"step": 2
},
{
"name": "format 2: payload truncated",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[80682, 1, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 2: payload age modified",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[80682, 1, "e9"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 2: control modified",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[594, 1, "50"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 2: non-canonical dk1_ JSON",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 1, "7c"],
[51, 1, "56"],
[74, 61, "4169626d56306432397961794936496d5268644756725a586c7a4f6e463161574e72626d56304f6e597849697767496e4a766457356b496a6f784d4441"],
[135, 0, "776651"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_DATEKEY_NON_CANONICAL",
"step": 4
},
{
"name": "format 2: unknown profile",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 1, "76"],
[51, 1, "50"],
[100, 32, "5a585a74626d56304f6e597849697769636d3931626d51694f6a45774d444239"],
[132, 3, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNKNOWN_PROFILE",
"step": 4
},
{
"name": "format 2: release of another round",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 2: access_policy=time_only with time_and_key structure",
"spec": true,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
[136, 1, "00"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250e3c7be83cbf1fbd6b115c96411b3bd010350fe68d3b3bb59db0c14b7ef57d8d6cb5e04667832353531390558209c09aeda759a8664a98120435fe0f6bf850000ac5ef0f9853ff46144c365780106a1005820600892659fe4890223e895876275f656995d170fda42b07fb2bec0ca51ce4b43",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 2: access_policy=time_and_key with time_only structure",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[136, 1, "01"]
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 2: extra stanza in OUTER_TIME_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 5
},
{
"name": "format 2: extra stanza in PAYLOAD_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 6
},
{
"name": "format 2: non-X25519 stanza in INNER_ACCESS_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 2: tlock stanza round differs from DateKey.round",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_ROUND_MISMATCH",
"step": 8
},
{
"name": "format 2: tlock stanza chain hash differs from the pinned profile",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_PROFILE_MISMATCH",
"step": 8
},
{
"name": "format 2: extension data of a type other than bstr",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "9f000001caa6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c0101026c7075626c6963206c6162656c"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "format 2: empty extension data (h'')",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[11, 6, "93000001caa6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c01010240"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "format 2: 65 extensions in one array",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[10, 7, "0590000001caa6"],
[137, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "format 2: tlock stanza U re-encoded with c0 + p",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[303, 64, "47767356342f426a37792b42394835566479466e705a316763796665646e334254332f2b66554158746e53706d30575865732b6e7633706e30632b4b6272676a"],
[416, 43, "58384841766f4779353462544e456f356a554a3456416f694e5459656555504745383370754a5654695a73"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 2: tlock stanza U is the point at infinity",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[238, 129, "774141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141410a41414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141"],
[416, 43, "62433146764a75425074647a376738755a595536322b68313332482b397966334a556c6639497455742b63"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 2: tlock stanza U with the infinity flag and a payload",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[238, 1, "78"],
[416, 43, "4a5574764474646b44706658326f393336744b625466784255453572536d4e54524f566b54777635667773"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 2: tlock stanza body of 127 bytes",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "c9"],
[409, 49, "410a2d2d2d20664d3147664471544772417532736948524e58472b75394e47766437313333564a6b5966724f3559736534"],
[458, 1, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 2: tlock stanza body of 129 bytes",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "cb"],
[411, 48, "410a2d2d2d20324a39614a58625751566464762f2b656d6d4d6c4b7a30546365706b6c344e35547966495635586a4775"],
[459, 0, "49"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 2: release signature re-encoded with x + p",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1004,
"signature": "be076aa25a40df5b4d22f9f7bcedaff30dcac20d94556107b8e652c7b54b2a440ce796f9fa53ad28023c84b40a580f55"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 2: release signature is the point at infinity",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "c00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 2: release signature with the infinity flag and a payload",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "d44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 2: release signature negated",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "944679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 2: negated release signature and U re-encoded with c0 + p",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[303, 64, "47767356342f426a37792b42394835566479466e705a316763796665646e334254332f2b66554158746e53706d30575865732b6e7633706e30632b4b6272676a"],
[416, 43, "58384841766f4779353462544e456f356a554a3456416f694e5459656555504745383370754a5654695a73"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "944679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
Format 3, step 3: the reader Open reads format 3 (spec 29.2 to 29.7, 63 steps 17 and 18): the PRELUDE accepts VERSION 3, and the files go to a Sink. - Sink: Begin with the validated head, Create for each file in the order of the head, and Commit only after every check of step 17; after any failure that follows a successful Begin, Abort, once. A format 3 capsule without a Sink fails right after step 2 with ErrSinkRequired, a caller error with no code, no failed step and no request; a capsule of format 1 or 2 without dst fails there too. - Step 17 in its substeps: the frame and the area, security and its verdicts, which never fail, the head, the files filling CONTENT, the SHA-256 of each file and the padding. A failure of age or a plaintext whose length is not P prevails; otherwise the first substep that fails decides, and a code other than ERR_INTEGRITY is reported only after reading PAYLOAD_AGE to its end. - The reads of BODY grow with the bytes received, never with AREA_LEN, HEAD_LEN or a declared size (spec 57); a test measures it. - A failure of the Sink is the caller's own error with ERR_INTEGRITY, as one of dst is in formats 1 and 2. - Opened gains Head, Verdicts, AreaLen and UnusableHeadExtensions. - Test data: "version changed" sets VERSION 4, and the format 2 list gains "format 2 time_only relabeled format 3", which fails at step 14, as section 64 of spec v0.10 lists: 126 cases, 89 of the spec. The randomly built capsules keep their recorded bytes. - testkit: Build writes format 3 and can edit the padded plaintext; Head3, Body3, DiscardSink and MemorySink build and open BODY. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
{
"name": "format 2 time_only relabeled format 3",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[4, 1, "03"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 3: the reader Open reads format 3 (spec 29.2 to 29.7, 63 steps 17 and 18): the PRELUDE accepts VERSION 3, and the files go to a Sink. - Sink: Begin with the validated head, Create for each file in the order of the head, and Commit only after every check of step 17; after any failure that follows a successful Begin, Abort, once. A format 3 capsule without a Sink fails right after step 2 with ErrSinkRequired, a caller error with no code, no failed step and no request; a capsule of format 1 or 2 without dst fails there too. - Step 17 in its substeps: the frame and the area, security and its verdicts, which never fail, the head, the files filling CONTENT, the SHA-256 of each file and the padding. A failure of age or a plaintext whose length is not P prevails; otherwise the first substep that fails decides, and a code other than ERR_INTEGRITY is reported only after reading PAYLOAD_AGE to its end. - The reads of BODY grow with the bytes received, never with AREA_LEN, HEAD_LEN or a declared size (spec 57); a test measures it. - A failure of the Sink is the caller's own error with ERR_INTEGRITY, as one of dst is in formats 1 and 2. - Opened gains Head, Verdicts, AreaLen and UnusableHeadExtensions. - Test data: "version changed" sets VERSION 4, and the format 2 list gains "format 2 time_only relabeled format 3", which fails at step 14, as section 64 of spec v0.10 lists: 126 cases, 89 of the spec. The randomly built capsules keep their recorded bytes. - testkit: Build writes format 3 and can edit the padded plaintext; Head3, Body3, DiscardSink and MemorySink build and open BODY. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
{
"name": "format 1 time_only relabeled format 2",
"spec": true,
"dkc": {
"base": "time_only.dkc",
"edits": [
[4, 1, "02"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
{
"name": "format 1 time_and_key with one stanza relabeled format 2, with the identity",
"spec": true,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": [
[4, 1, "02"]
]
},
"identities": [
"AGE-SECRET-KEY-1840Y650JZGMWEL6QMZWQ4H8E4GE95F9ERDR0M8G6LFYZ749J6TRQR3F5PZ"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 2 time_only relabeled format 1",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[4, 1, "01"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
{
"name": "format 2 time_and_key relabeled format 1, with the identity",
"spec": true,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
[4, 1, "01"]
]
},
"identities": [
"AGE-SECRET-KEY-1NSY6AKN4N2RXF2VPYPP4LC8KH7ZSQQ9VTMC0NPFL73S5FSM90QQSN6LZCL"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
{
"name": "INNER_ACCESS_AGE with 15 stanzas",
"spec": true,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[14, 2, "07ee"],
[900, 1267, "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"],
[2167, 98, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"identities": [
"AGE-SECRET-KEY-1NSY6AKN4N2RXF2VPYPP4LC8KH7ZSQQ9VTMC0NPFL73S5FSM90QQSN6LZCL"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "INNER_ACCESS_AGE with 17 stanzas",
"spec": true,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "b2"],
[2067, 198, "5b8d2a9254bf06b0f4ceaea6fc789f29893ceb0a496e2b856a8fa67abaaeef8522521584d2072d595d3502c6b41aab40ca166ee2c5bbc08497ac1c0e99cfc89c3da0f6e12d206de37c00f3bd0b2ef94a78e60f19cdd07f1653f920dde0330042a25f0634fa4371dcfc680a195229823aa217216b9bc5e583fe75f5edaf0512635610fd5d85a5ce57b9e8d7c1962eebd8070cd796bc2bc7c882065dcfd80fa821c9ed3c523acc778d111a16f2a50f8d8d2cddd85f4412b495d63f6fdf6e02fa667cc374158c30"],
[2265, 0, "1378b1dc03577d28aed9e20f8f9916e6205124248ea6d643a862e5d2ce3b73a0630cd8737d91a2f5477ecef216b7af503597ef9e4d90103c92a63cca2adc179d64cd1831cb21c20f29dccb4c5cd66989ebf2a2236a45d7eb6dc969eba16c3d60b366"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"identities": [
"AGE-SECRET-KEY-1NSY6AKN4N2RXF2VPYPP4LC8KH7ZSQQ9VTMC0NPFL73S5FSM90QQSN6LZCL"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "16 stanzas, two for one recipient, and the identity of that recipient",
"spec": true,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[900, 87, "9394dcd109377c8ebbd67040d34d547b1250e330946719f3280ea614dcf7004f8c30d5202dd43c6bbd767533b19aee2f5af5fe6b77108f03f5272a755557e63d332e38b4d1a5388b162f7358510418b83ccb754a457c68"],
[2070, 43, "911bd346a690ce82ffa76280299935d25355060ce86cbf8621d480a09f206f36a9b4173c5672012ec4b035"],
[2249, 16, "4c220e9720a284e83667a897f499e3de"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"identities": [
"AGE-SECRET-KEY-1NSY6AKN4N2RXF2VPYPP4LC8KH7ZSQQ9VTMC0NPFL73S5FSM90QQSN6LZCL"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 13
},
{
"name": "identity that is not a recipient of any of the 16",
"spec": true,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": []
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 13
},
{
"name": "control of schema version 2 without key 6",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "c0"],
[476, 1, "a3"],
[500, 32, "01334cf961a9394f0b6d8b8fec6d566b2a2591b86da70351123a193e066db96e"],
[567, 17, "e81a6b2281478f480de71ef03342c64ed3"],
[584, 10, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "control of schema version 2 without key 7",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "c8"],
[476, 1, "a3"],
[500, 32, "e0b1747f1549994e6c7ae03391a26bffcbd83462fd2edee17b00aa674719dedc"],
[577, 16, "30cfec7a7d656b447630b4b439dfbabe"],
[593, 2, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "padding code 0",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[578, 17, "0d44858313fb885815265d4e1985d5390b"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "padding code 3",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[578, 17, "0e47dc0b8e0d25f76140bbcaf1b3956b2e"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "payload_length L_MAX + 1",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[570, 25, "83ab6c48c335f6580f0eccbf15b3e4a5da1a4fb320b4f992e4"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "payload_length of 7 bytes",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "c9"],
[500, 32, "5d3791fd50fb59de8a24ab1d59e3840615f9c9dcc8007010aae648150b53c721"],
[568, 26, "5f2f9c6b6c49f385f05d82e6fdf4fbf358c5e41fbcad20071df2"],
[594, 1, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "payload_length of 9 bytes",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "cb"],
[500, 32, "955d41376dd03b116ef0e8dc1ce9385256f62f550eea5ede5e70e64f25b58bbb"],
[568, 27, "512f9c6b6c48c334c7ef0abf53239233b3b90f0f62af5ff21e33e1"],
[595, 0, "2c"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "payload_length as an unsigned integer",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "c6"],
[500, 32, "ae9f92629835c3ea6cfff3763049b759020b68d3a636c03d123d4ef037f64fd6"],
[568, 23, "022f9d5bdc4fc1b9383c0700e591604b642c97143d390e"],
[591, 4, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 14
},
{
"name": "last padding byte not zero",
"spec": true,
"dkc": {
"base": "format2_time_only_extensions.dkc",
"edits": [
[1108, 17, "a9c4c81d0051c53edf9b90a846b2fd5e9f"]
]
},
"release": {
"round": 2000,
"signature": "b6cb8f482a0b15d45936a4c4ea08e98a087e71787caee3f4d07a8a9843b1bc5423c6b3c22f446488b3137eaca799c77e"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T16:49:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "payload plaintext of P - 1 bytes",
"spec": true,
"dkc": {
"base": "format2_time_only_extensions.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[1108, 16, "e0efdea667865b5fdf9dd98e76c33d25"],
[1124, 1, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 2000,
"signature": "b6cb8f482a0b15d45936a4c4ea08e98a087e71787caee3f4d07a8a9843b1bc5423c6b3c22f446488b3137eaca799c77e"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T16:49:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "payload plaintext of P + 256 bytes",
"spec": true,
"dkc": {
"base": "format2_time_only_extensions.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[1109, 16, "1a1d49d40c97b151f361236ad3a665cf"],
[1125, 0, "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"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 2000,
"signature": "b6cb8f482a0b15d45936a4c4ea08e98a087e71787caee3f4d07a8a9843b1bc5423c6b3c22f446488b3137eaca799c77e"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T16:49:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "payload plaintext without padding, of L bytes",
"spec": true,
"dkc": {
"base": "format2_time_only_extensions.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[887, 16, "d6ce427933325eeabe9d771bfc4a64ec"],
[903, 222, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 2000,
"signature": "b6cb8f482a0b15d45936a4c4ea08e98a087e71787caee3f4d07a8a9843b1bc5423c6b3c22f446488b3137eaca799c77e"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T16:49:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "padding code 2 changed to 1, with L = 78000",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[578, 17, "0c3c2efb98e8ecb9c80bffd140561508e8"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "payload_length L - 1, the last byte of the content not zero",
"spec": true,
"dkc": {
"base": "format2_time_only.dkc",
"edits": [
[576, 19, "58580f5c56ce65a8d754ee9c79638d30ac1df1"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 1 time_and_key with one stanza relabeled format 2, with the .dkk",
"spec": false,
"dkc": {
"base": "time_and_key_portable.dkc",
"edits": [
[4, 1, "02"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b6579010102503955e944a3c60cfa1fd6485e9693c77d0350448e134a13457c319cab7fceaf7ffa1f04667832353531390558203d5e4d51f21236ecff40d89c0adcf9aa325a24b91b46fd9d1afa482f54b2d2c606a10058202e97878078bae6358037a9c264f379a3cbe839f767d69836b0343f35657b2972",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
},
{
"name": "format 2 time_and_key relabeled format 1, with the .dkk",
"spec": false,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
[4, 1, "01"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250e3c7be83cbf1fbd6b115c96411b3bd010350fe68d3b3bb59db0c14b7ef57d8d6cb5e04667832353531390558209c09aeda759a8664a98120435fe0f6bf850000ac5ef0f9853ff46144c365780106a1005820600892659fe4890223e895876275f656995d170fda42b07fb2bec0ca51ce4b43",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
},
{
"name": "INNER_ACCESS_AGE with 15 stanzas, with the .dkk",
"spec": false,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[14, 2, "07ee"],
[900, 1267, "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"],
[2167, 98, ""]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250e3c7be83cbf1fbd6b115c96411b3bd010350fe68d3b3bb59db0c14b7ef57d8d6cb5e04667832353531390558209c09aeda759a8664a98120435fe0f6bf850000ac5ef0f9853ff46144c365780106a1005820600892659fe4890223e895876275f656995d170fda42b07fb2bec0ca51ce4b43",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
},
{
"name": "INNER_ACCESS_AGE with 17 stanzas, with the .dkk",
"spec": false,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[15, 1, "b2"],
[2067, 198, "5b8d2a9254bf06b0f4ceaea6fc789f29893ceb0a496e2b856a8fa67abaaeef8522521584d2072d595d3502c6b41aab40ca166ee2c5bbc08497ac1c0e99cfc89c3da0f6e12d206de37c00f3bd0b2ef94a78e60f19cdd07f1653f920dde0330042a25f0634fa4371dcfc680a195229823aa217216b9bc5e583fe75f5edaf0512635610fd5d85a5ce57b9e8d7c1962eebd8070cd796bc2bc7c882065dcfd80fa821c9ed3c523acc778d111a16f2a50f8d8d2cddd85f4412b495d63f6fdf6e02fa667cc374158c30"],
[2265, 0, "1378b1dc03577d28aed9e20f8f9916e6205124248ea6d643a862e5d2ce3b73a0630cd8737d91a2f5477ecef216b7af503597ef9e4d90103c92a63cca2adc179d64cd1831cb21c20f29dccb4c5cd66989ebf2a2236a45d7eb6dc969eba16c3d60b366"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250e3c7be83cbf1fbd6b115c96411b3bd010350fe68d3b3bb59db0c14b7ef57d8d6cb5e04667832353531390558209c09aeda759a8664a98120435fe0f6bf850000ac5ef0f9853ff46144c365780106a1005820600892659fe4890223e895876275f656995d170fda42b07fb2bec0ca51ce4b43",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
},
{
"name": "16 stanzas, two for one recipient, with the .dkk",
"spec": false,
"dkc": {
"base": "format2_time_and_key_portable.dkc",
"edits": [
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
[900, 87, "9394dcd109377c8ebbd67040d34d547b1250e330946719f3280ea614dcf7004f8c30d5202dd43c6bbd767533b19aee2f5af5fe6b77108f03f5272a755557e63d332e38b4d1a5388b162f7358510418b83ccb754a457c68"],
[2070, 43, "911bd346a690ce82ffa76280299935d25355060ce86cbf8621d480a09f206f36a9b4173c5672012ec4b035"],
[2249, 16, "4c220e9720a284e83667a897f499e3de"]
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250e3c7be83cbf1fbd6b115c96411b3bd010350fe68d3b3bb59db0c14b7ef57d8d6cb5e04667832353531390558209c09aeda759a8664a98120435fe0f6bf850000ac5ef0f9853ff46144c365780106a1005820600892659fe4890223e895876275f656995d170fda42b07fb2bec0ca51ce4b43",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Implement capsule format 2 of spec v0.9 The reference moves to the DateKeys Protocol Specification v0.9, approved by its author on 29 September 2026. Encrypt writes capsule format 2 only; Open and Inspect read formats 1 and 2, and a format 1 capsule keeps the verdict v0.8.2 gave it. Format 2 (spec §22, §29.1, §31, §39): - VERSION in the PRELUDE is the capsule format, capsule.Format; any other value is ERR_UNSUPPORTED_VERSION at step 2. - CONTROL_CBOR has the schema version of its format. Version 2 adds key 6, payload_length (8 bytes, big-endian, at most L_MAX = 2^53 - 2^46), and key 7, padding (1 bloque256, 2 reforzado); it is 103 bytes without extensions, whatever L. - The payload is the content padded with zeros to P = rule(L). Step 17 checks the length and the zeros, and Open writes only the first L bytes. - INNER_ACCESS_AGE holds exactly 16 X25519 stanzas: 1 to 16 credentials, and a dummy in each slot left, in a uniformly random order. Writer rules (spec §62.1): EncryptOptions.Length is required and the source must deliver exactly that many bytes; recipients that are not canonical or of low order are rejected (agewrap.CheckX25519Recipient); self-checks of the header, the control, INNER_ACCESS_AGE and PAYLOAD_AGE. The CLI measures its input, takes -padding and reports the format. Test data: seven format 2 fixtures, padding vectors checked against math/big, format 2 CBOR vectors, and the mutation corpus in both formats with the 22 cases of the third list of spec §64, built without randomness by sealing the fixtures again with their known keys and nonces. The format 1 fixtures are kept byte for byte and never regenerated; the differential corpus keeps its 1825 cases and adds a block per format 2 fixture. The spec copy loses its "to be implemented" markers, and the READMEs, CHANGELOG, traceability and testdata/README.md follow v0.9. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
},
{
"name": "format 3: PUBLIC_HEADER_A + SEALED_CONTROL_B",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_HEADER_BINDING",
"step": 15
},
{
"name": "format 3: SEALED_CONTROL_A + PAYLOAD_AGE_B",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 3: DateKey A + release of round B",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": []
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_ROUND_MISMATCH",
"step": 10
},
{
"name": "format 3: chain hash changed",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[173, 64, "61626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162616261626162"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_PROFILE_MISMATCH",
"step": 8
},
{
"name": "format 3: version changed",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[4, 1, "04"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 2
},
{
"name": "format 3: flags != 0",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[5, 1, "80"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_FLAGS",
"step": 2
},
{
"name": "format 3: reserved != 0",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[7, 1, "01"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_INVALID_FLAGS",
"step": 2
},
{
"name": "format 3: payload truncated",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1562, 1, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 3: payload age modified",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1562, 1, "55"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "format 3: control modified",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[594, 1, "80"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 3: non-canonical dk1_ JSON",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[11, 1, "7c"],
[51, 1, "56"],
[74, 61, "4169626d56306432397961794936496d5268644756725a586c7a4f6e463161574e72626d56304f6e597849697767496e4a766457356b496a6f784d4441"],
[135, 0, "776651"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_DATEKEY_NON_CANONICAL",
"step": 4
},
{
"name": "format 3: unknown profile",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[11, 1, "76"],
[51, 1, "50"],
[100, 32, "5a585a74626d56304f6e597849697769636d3931626d51694f6a45774d444239"],
[132, 3, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_UNKNOWN_PROFILE",
"step": 4
},
{
"name": "format 3: release of another round",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 3: access_policy=time_only with time_and_key structure",
"spec": true,
"dkc": {
"base": "format3_time_and_key_portable.dkc",
"edits": [
[136, 1, "00"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250bdb483fba42daf0b409f44d23033f362035029737cb54ad5310734ecf20720c2d317046678323535313905582042e50b6ae3b2f00b79f08609f7fe11fba1256c54c6faa0c9ec9258b5f6f4c05906a1005820680d29962e575689a31543df28433dae7737abd9a793e9cae92ef40920d09636",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 3: access_policy=time_and_key with time_only structure",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[136, 1, "01"]
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 3: extra stanza in OUTER_TIME_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 5
},
{
"name": "format 3: extra stanza in PAYLOAD_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 6
},
{
"name": "format 3: non-X25519 stanza in INNER_ACCESS_AGE",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"identities": [
"AGE-SECRET-KEY-1G2PR0YP2Q6DHNQV9QPC7PFAC34AGUUU9CGJGJQCLJR3QLAUYPMGQ2LKA30"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_POLICY_STRUCTURE_MISMATCH",
"step": 12
},
{
"name": "format 3: tlock stanza round differs from DateKey.round",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_ROUND_MISMATCH",
"step": 8
},
{
"name": "format 3: tlock stanza chain hash differs from the pinned profile",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": false,
"frozen": true,
"error": "ERR_PROFILE_MISMATCH",
"step": 8
},
{
"name": "format 3: extension data of a type other than bstr",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[11, 6, "9f000001caa6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c0101026c7075626c6963206c6162656c"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "format 3: empty extension data (h'')",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[11, 6, "93000001caa6"],
[137, 0, "0681a300716f72672e6578616d706c652e6c6162656c01010240"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "format 3: 65 extensions in one array",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[10, 7, "0590000001caa6"],
[137, 0, "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"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 4
},
{
"name": "format 3: tlock stanza U re-encoded with c0 + p",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[303, 64, "4d6151686359416d2f674f615757554a3175326259484f7876665368707976737572364c327151656a715a654b374c652b474a672b4f476d59355552544b7772"],
[416, 43, "5272496b5a574c4e495975574d3535336c347a58764a77785662327a595675794a436b2f6b304f78374f63"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 3: tlock stanza U is the point at infinity",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[238, 128, "774141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141410a414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141414141"],
[416, 43, "634174475343333330735a2f5777484f61692b4970722f414d6938345354424e31344f2b312f467a477a45"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 3: tlock stanza U with the infinity flag and a payload",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[238, 1, "78"],
[416, 43, "3359707a43436e62426e616e77714846326a5153654f6a453350653848734567424e4d58504e645453576b"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 3: tlock stanza body of 127 bytes",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[15, 1, "c9"],
[409, 48, "410a2d2d2d202b78684c412b647945726a4f596f62376e5a304f67657773744e565a3471736d6b6b6763595463487a30"],
[457, 1, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 3: tlock stanza body of 129 bytes",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[15, 1, "cb"],
[411, 48, "410a2d2d2d20654f44547a4736304e483948744c4237324441694946466876536a444d5976434957714359324c6f5638"],
[459, 0, "45"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 11
},
{
"name": "format 3: release signature re-encoded with x + p",
"spec": true,
"dkc": {
"edits": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
[0, 0, "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"]
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1004,
"signature": "be076aa25a40df5b4d22f9f7bcedaff30dcac20d94556107b8e652c7b54b2a440ce796f9fa53ad28023c84b40a580f55"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2024-08-23T15:09:27Z",
"registry": "default",
"network": true,
"frozen": true,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 3: release signature is the point at infinity",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "c00000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 3: release signature with the infinity flag and a payload",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "d44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 3: release signature negated",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "944679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 3: negated release signature and U re-encoded with c0 + p",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[303, 64, "4d6151686359416d2f674f615757554a3175326259484f7876665368707976737572364c327151656a715a654b374c652b474a672b4f476d59355552544b7772"],
[416, 43, "5272496b5a574c4e495975574d3535336c347a58764a77785662327a595675794a436b2f6b304f78374f63"]
]
},
"release": {
"round": 1000,
"signature": "944679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_INVALID",
"step": 10
},
{
"name": "format 3 time_only relabeled format 2",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[4, 1, "02"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
{
"name": "AREA_LEN 0",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[781, 1, "60"],
[1547, 16, "0e7b699ff03465154ab884554c7521eb"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "AREA_LEN 511",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[781, 2, "6107"],
[1547, 16, "11de7e89c7381a8835af4fd3c301b50d"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "AREA_LEN 513",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[782, 1, "f9"],
[1547, 16, "923a3fc2536742ecd9faba3017fe9b1e"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "AREA_LEN 66048",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[780, 1, "85"],
[1547, 16, "1d1eff28283a2b515d8c3eb9a90ceaa0"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "SECURITY_LEN 0",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[786, 1, "a3"],
[1547, 16, "6aa4a70020cb3b7d6ceddabe54eb09c2"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "SECURITY_LEN 513, larger than AREA_LEN",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[785, 2, "a5a2"],
[1547, 16, "6b794e809f0b665a09bbfde7c4a8d3e6"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "HEAD_LEN 0",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[790, 1, "b9"],
[1547, 16, "e74aa70659c537e6ae84b8d35213d3ca"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "HEAD_LEN 2^24 + 1",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[787, 4, "6e1f0db8"],
[1547, 16, "0758b8663319b8196ad57e6c865ccdbd"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "12 + AREA_LEN + HEAD_LEN = L + 1",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[790, 1, "31"],
[1547, 16, "9d2d1c486f5fffd2a0e92755d1fe7ab5"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "L \u003c 12: 11",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[575, 20, "6dda8c2ebe78dcad444f9e3f035bccb0f804905f"],
[790, 23, "b9e68be55a8ce70633c01029070ba951ba9f20cb6dfa1d"],
[1035, 16, "0407d9577dc7dbfc066dd4e95c350422"],
[1051, 512, ""]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "a byte of the area not zero",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1302, 1, "a7"],
[1547, 16, "83277cf7b483cbf2e30a194abcd69489"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "head of version 2",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1320, 1, "c7"],
[1547, 16, "24ab7d4091dd239b4747064ba45a5ea8"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 17
},
{
"name": "head of another type tag",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1318, 1, "da"],
[1547, 16, "5d65346cdadabe720d238754b71bc700"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 17
},
{
"name": "head with a byte more within HEAD_LEN",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "458c2ed630a93baac2248b4af35b4f67d18c2e"],
[790, 1, "cb"],
[1416, 23, "a77044c81f0334c1ae3d12ae95bf912fa55fa15ac8e226"],
[1547, 16, "6f4a754f9340f32131be4a6bb599b939"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 17
},
{
"name": "paths b and a, in that order",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "628c2e6abf142176591038e4b59d14cad3d25f"],
[790, 1, "28"],
[1357, 113, "dc71d99ae1ac41235a5e5922e9501e3d6d67e87c67f88e6b824800d775bd3bd9cf554efb82ba951700c3b02c853159e8d021785d21a77ec9214dbf4488efe6e6df4cb047a2834669bc087be38281dac357bff6db37116943b4027db89d610957639f3227ba7f84c79b99b366db5a168992"],
[1547, 16, "badec288bbbcd42507e88a68d1b58d3f"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 17
},
{
"name": "paths b/.. and a, in that order",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "678c2eef415cf0aa35c6f48b38a0363c9f862a"],
[790, 1, "2d"],
[1357, 116, "dc71d99fe182790f5b4b4d26b266473c2065f8d04141a3210de223bd6c6f19612a9b77342813d8b8911c704fca1d79dda1e5793c42b07ddd331689a3600b47cee7123c214563140b2779c6ab95e419098dc8dee431c4d6ae7fa150a24a5820186b9b6127ba2c85ce9b8cbe35ca5f0ac6fcbb294e"],
[1547, 16, "f8db3ddefe83414abd8add3f0202fe0d"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 17
},
{
"name": "path of 1025 bytes",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[575, 20, "6b5f8c2ee837777ad9430ca7b0981ad8e2c9c4b0"],
[789, 2, "09d5"],
[1360, 203, "8287ac360e3b722e09d05f224b6117c0927f2703779c2171ed4b506d118983f13cb0e251c48796ed86d10e9dc7d06f1872429e1cf0443afe885904c551427fc5f23858efdfb29e3eeb4def5f88ad03db50c3a71314879549df9322095866d5206cbe708084dac6b33adb111888f9fb666b6a887192fb71616dd84b38b02b3ab6948ebb7fa46b17eae433b9fcabc832bc3b67ae1091ace50be781343b4a0f5f8fa2a3311f1543b92de16d889513c2142cc4d45246ee1dd9a44f9d0913c4ae4bdc97a4d2b7196f7d7468b715"],
[1563, 0, "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"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_NON_CANONICAL_CBOR",
"step": 17
},
{
"name": "path ..",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "5c8c2e6abc8b5bd0ea02bce7e5cdeb0d462e06"],
[790, 1, "d2"],
[1360, 78, "99ad835637585d4c30b52863395955fef4df7192b4ce7156d23f20ce090f8b69cd399e474e1a5ff6953514ecedab2ac5a4e3f912b34435fbc24b4fc15e463c84ad3804ef94bc9f55c42cc03ea7cc"],
[1547, 16, "08aeb56ea43ee3d109e9dba78a01f846"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path /a",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "5c8c2e6abc8b5bd0ea02bce7e5cdeb0d462e06"],
[790, 1, "d2"],
[1360, 78, "99accc5637585d4c30b52863395955fef4df7192b4ce7156d23f20ce090f8b69cd399e474e1a5ff6953514ecedab2ac5a4e3f912b34435fbc24b4fc15e463c84ad3804ef94bc9f55c42cc03ea7cc"],
[1547, 16, "e658224eaced5002b1570c08c84364d0"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "paths A.txt and a.txt",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "6a8c2e74c320d3fdeccc328a20084b1ae658d7"],
[790, 1, "20"],
[1357, 121, "dc71d99ec28323592e5c5924b17355605818fce473572b9984fefcf155391c2bb76ea7f4f634ee03b5ce2955565bc06d8450dc5d46d053ab5d619ea73a3da7682708843ee9b3cc682fad4b5ed734f63ea3959e580cc9965f7d2964546794730f5542140bb033808bdf8ca171df1e1ccbb8a466376ac37f9390"],
[1547, 16, "ea335643eca779e48dcb03641e2ccb76"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "paths ab and a, U+200C, b",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "698c2eef42dfdc8c5a757335d33a44d0236848"],
[790, 1, "2f"],
[1357, 118, "dc71d999e2cf5637585d4c30b52863395955fef4df7192b4ce7156d23f20ce090f8b69cd399e474e1a5ff6953514ec4db12518bfa33d1fde251789a42e2ffc5ec0e0609fc18b92e4494aab0cb5af8783eb82fb9bc613e177422fb1eb8a5fd0224f952d22ff3b84d8df8cf270d61e09c6ebb5632b25ad"],
[1547, 16, "10f450c1498c5ac761ef57248baf32fa"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path CON.txt",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "438c2ed62f264fc89d750ca158c141d34cab10"],
[790, 1, "c9"],
[1360, 78, "9cc0e2190f2e253b27a7724367163cf99d43510f07b4cf6932c64c7378f8b9cfd1e62a17123851dfde15dd1a928dc54c4d26ab1763dcd5d7c8544a841a4623c0b87912e2d0a3d02ca548af10adcc"],
[1547, 16, "144f8b910fa221ecde7bf850712865d8"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path CON.txt in full-width forms",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "498c2edb32af146e0c83889d28916a8fda4f6a"],
[790, 1, "cf"],
[1360, 83, "966c11f4cee6f2a09a1f5e371c7439b7bf1e0b745ca52e4d9b472045ff915c7c20ae8b20feb97bbb4e97d82200e237dca1c5cb415a10059abd52975dbe6b3fc8bc7913eb83b7d47fa140e04ec6bf4dde108c82"],
[1547, 16, "34876e09d8254e37d0143f2951a25cbd"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path ABCDEF~1",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1361, 8, "c2ef14651f1b3117"],
[1547, 16, "c5a8b5f82723403bc00ebcbf583c5327"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path with U+202E",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "458c2ed630a93baac2248b4af35b4f67d18c2e"],
[790, 1, "cb"],
[1360, 79, "92e24fd78f38733b5ec5715566003bb7b946283f0190515903538fe00d97ea80068e0f070cc1405241e3565914573a45d816a178c54fa966677044c81f0334c1ae3d12ae95bf912fa55fa15ac8e226"],
[1547, 16, "bce91df939b5778cd233e1dba70365ab"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path .datekeys-x",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "478c2edb312c288ce7d309f48df65cfb556e4c"],
[790, 1, "cd"],
[1360, 81, "90adc936553f362a5fc25d3b65163aa1be080c3a78a0577d9d63be75ce049fef55c1d86f29d15eab506ec9af9f14bc9f271f3448cf200ff51bc1ebec114f3184b93c04ea95f3d433e45ca14dc6a8439475"],
[1547, 16, "f0ed94acc6218f9de2ebe6ade9a0f8c0"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "comment with U+202E",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "4b8c2edb3332015031320747c32b78235f3188"],
[790, 1, "c1"],
[1303, 1, "03"],
[1356, 89, "d03bb53b7b2dcf52a0fc5d2748de04224a7440d5bc080a625beb0a48e277266161a16de9e33dfe4fadf6acd36b87c6db116fa82a2a864d9ba51990aab73d31cec2da90ecb23516ae94b6c23ba10ca55287bc4dc91ec6e75c31"],
[1547, 16, "c31f0d95fb578d3440e26030140083cc"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path a.",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "5c8c2e6abc8b5bd0ea02bce7e5cdeb0d462e06"],
[790, 1, "d2"],
[1360, 78, "99e2835637585d4c30b52863395955fef4df7192b4ce7156d23f20ce090f8b69cd399e474e1a5ff6953514ecedab2ac5a4e3f912b34435fbc24b4fc15e463c84ad3804ef94bc9f55c42cc03ea7cc"],
[1547, 16, "518175e072e0515779cdf8fa88d0948a"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "paths A and a/b",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "648c2e6fc0970d587ebfb68d5038225e58b47d"],
[790, 1, "2a"],
[1357, 115, "dc71d99ac2ac41235a5e5922e9501e3d6d67e87c67f88e6b824800d775bd3bd9cf554efb82ba951700c3b02c853159e8d221563f22b17fc926039bff18c814ba61c858c14ded75041b28307d08e790e6284821ecee2ce7a2fdc27a48a745071b66da2526ac3b848bde85f265db4d18c3f7fa0d"],
[1547, 16, "3591b9289ad15a88635ffef7cbcf465d"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "65536 implicit folders",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[574, 21, "20fafc8c2edabadee2e6bfa46e749ff116fec042dc"],
[779, 784, "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"],
[1563, 0, "a9daef49219f101de2b18e382eaf8785a5be681467f78c32e6905d5e2c9bb096f4a87010e5608303c243cc57027bca9773f453949b8f662e21957967dd686046eacaf88fbb2da20a2bbc4e41a06e7c19ab8dfa5ebe6291ae8a07beb08d2a1ca4c57b8626fc8a1dcaabb7367f4596843ae4bd5a1a1dcfc41b189ff13b2186e9de7050018cc724f8d5cc008afe16b0e2ec51eb38119862e0966d2ca84ea32173f3ce8e327521c87b4567c98a16787f7bef471026dd19e1dfe2561a2268b082ba094c047ef3768d3cf0c08944c62fb3f2ef793b77f54036605658e2093383a2bc5619d38565313fbdeef7d7c406713a3f5f65f449c8dfc890931519ed0d7ee0b9caaa12a7e8313799bc61ef96eaf9ac3f5278b12789e4ab064fe61a4ebc9e5db52cafb31aea86f9a064e652a0007181799db963056b6d899ecaee7a000a69e27db41d08e641ad7c645bd6b12e7100958b6740d0cd4b512130745157df198d1d5674cc544cce0d4763b2b21cfecaf1092306ab6e9accc470aee61e168a0a90933bece368258834ca311d13ea933fc64176f26706f74eff404886429c7ea1a33f4a9b3b994f08f29a1f74ac4c54ef108c27771cf4a71a88b8ea85617275de7d3075d16da47a93a39af79e594de4931e0d5ec3914f562cb3ce8e5a64165a2bd57ba469b6b5dba4155baa42b0d35af2b47ee344a84942cdc5e108fa6f4f301ac94a11bae9f488122e4f3f12a22a2cfd897fe65d08bebf206eb727477a4cf45d642e2b68eebca3b16d7a9106cd240e3816fe00442f775747053575a950d0f235425ea740a9c10a7a7306049f5b98c669442d3bdea9a62643779a49b84aaf3104d5556690c8df6af6f13aa6b8a9719689e3a27ba3beac8b1769b786804ad3dc793c0fb3ad5e485e96304aaf7835553eb367e6edc37d5e475fc280c5ad7508a0337b96277a283679cd93f62bec17e8161c1186fab6460d5a1a63acf11820c259f904814992823103b69a979ba26c2e34d22b0e4bf47c3a10e2909a60ed843491f2aba9f84687d874db6db1e2cc6178c10d4ce7fd42299a01cbd2372e27595be398a80cbccff005a8684f4f86159181f21dabc56c34da3e3b002e6d64ba546c955f80fd89e37dc36c4e5f05376139d58403e4ade3cc8ba306acfc8ff5dcaf3b76265bebf7f8fea9298747583c4561f1bbec405720090ee6258f14d321d54abe67290f49ec5c8bc12fcfaf581c4165f8a00fdd110c1f51dd50fdf0f6d7b5100ad3b99fc487a91257c503cf8830a9589f4a9ad5a853a65f892700ad4969d0c4123165b855c1b45e2f6f44506e7e87ca684979a60cc62fd00b5714df24a766a62893422ba74ea26b9c1d32916500fec98df28a5cb999b7cde3ab018cf28415a2830bb9fb9ae85c6127dc6d1ca3794eed23fc27a896b2c40cd03e881ca23333e518d0650578ef3977163da18bea808c4837b50ea2aeb90e1224d35ef4609b00d194179e0def9390b8aec2f588f02eff356bbedf355eae605f211d488a3eba182163d583f0310bb1b3b5f3a87691d83df8e866533d58d99a18922ee9604ee41688ee8ce4ad3a0d3a4180e20ac06d6220f3b0275b410d424cb32b303f51c85ecedd538be097e7561267b03d237c42034ce0135933d583d0f13eeee082a2bb701387f53c71587c9121713e8f8cc96082c905d12d1c56c2c0996f7d659449a47e7dbedaea9408303d9c4f0c9cb351ac9f089837349a5d916b26ca89094c03750e27fa263794052f2cb452ce7259ed3b2a1b2f9266fcbfa4a2ea236e978659a94ff2dfff0983a318b70ba40fabcfeaf509a4c6b20f467e9be49a314baf60b07b6cade8348abda1232db64d3e15f9c33be7b393e12b3eb4051c3e708b8680571969a36015cd245cc12e88426b78e9413ef619cf108450e4532281a72527e3db5c746461001f17c5600a5c769d9be63416656c06cf8e942218d171d10051f4388de0e642900286fa2bb391d2f48d4743fbc4953298a2193cc4b25a2ddcdeff8ff62b32d00c114902d16455a2e68351c2f62e24eef1a13a69029afc2602cbcd0435094ca7f705a633c800ec97c0b8c15863f6c4d1a47d1e57ae5b619462c8ac3f5030ff8854e6e8ae67467f7b581fb4a50dc390e7758945e03614254eb823e16ff55cdb52eb295020d5b04dfc8f904769fbc7bfb57c87cf2b5d135205ddb1e6e4a235050f13c7f3bbf0fea937567e19853f636b9826d003afe7f3eb6ba4a1fe0aa07d02633e4cffb564e1527fc987f405dfc21079ee73cb828cfe4cd3c938e8c072928e6dd30c3a3b94473350702ba4db9a83399dd360095cb67b89745b322819e13605289a94bb022df1a0c398d8d37f6440f32d865f5ddefbe5d00bba4c9eb4f25a35e2dc9a6cc5db15bd7ea6989b5cd58cd4ccfe4ae55432902b0324f15707dab90fa2ccfda6c2e3f0284ee9457b331f0035bac5b24e590771edd3fa32ad3abb7da49b734554ca56d46633eb5280ef943de0500228eab8825cce74f531fc92a856b69dc7fc823a92f6c57b7227cc2999a187441f51dc43a215ea5249e6d7f8cf747f719b4a2e1698a642e94bd3cea068b14dadb1b7c20c23ef4db93101aba75d7c6e75fe5ec65b0a4dd0818042875943d0b4b69f7aa6ab9458ef53bae686cb28194e650f0ecc6181844ae34bd165b546c7e8c1d6f09444ea6e7d02d306bd56d88eb0504defd044328e2b4b72f1e5a395a48157a1aef160bbe3ff888b08e8f123eb3eb01433661634e0fe1e63c6d66b4f2e9e04474b9de63ac8fbfbac04ea93c67706b3246cbffd0f16316cf4a30b04843930ecbe6c91c21f9339065bf2c6311a9d5b2b39d41b7a70a111cfc15f20b3c9104776d7c65b8ca730cf7cb8850e96
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "declared author with LF",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "4e8c2e60b5f6e3a2e8384545ab938c01a60335"],
[790, 1, "c4"],
[1303, 1, "03"],
[1356, 94, "d73495b79a89e194922a38352330d6430c6e57d5dc307c1a2cfc1812c229690866c8f1c97e8e84f1921655bfd6f631e9b773779e7ada6f958c52b063414217214b3355beb7e58e4eb8bcdd3ee448a54dc3a90cdf1382f8134887de47def8"],
[1547, 16, "191833edc33401512aa3a574645c353a"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "start of the first entry not 0",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1372, 3, "65032f"],
[1547, 16, "2b3e70b9e9128f6d4e36774fd3eb4e75"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "end - start not size",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1374, 1, "2d"],
[1547, 16, "5b378380ae42cd55ecf5c884576a1084"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path a followed by VS16",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "5e8c2e6abd0e48b20fb23a918068f9a1ca0f24"],
[790, 1, "d4"],
[1360, 78, "9fe242efae5b4b4d26b266473c2065f8d04141a3210de223bd6c6f19612a9b77342813d8b8911c704fca1d79dda14563379f48bd974a79fe875c4ed71a4670c1b17907ef83b2d530ea26c03ea7cc"],
[1547, 16, "2b45360d3ea41c8e93642cc720b0d8e4"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path with two ZWJ in a row",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1361, 8, "e24fd7acb8ddc244"],
[1547, 16, "600cf553b447875fc5a67a1a93b69b92"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "comment with the tag U+E0041",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "4c8c2e5bb473f7c0c389c69b10f97e6d212217"],
[790, 1, "c2"],
[1303, 1, "03"],
[1356, 92, "d0369cb697e25ef7a0db58ce80b1182d0b74598fc9667c634eff0e13d42e272c63b1c1cf5a10b4c007d5c6cab9a57e3edf56678083cbe20a7ad9f3e59bbd1eb47b393a1824993fe19cb2913ba15fa45b87a9409a0fc3fb135f899422"],
[1547, 16, "b2431862112dcfcb1e093cbc5cbc6877"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "comment with the variation selector VS17",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "4c8c2e5bb473f7c0c389c69b10f97e6d212217"],
[790, 1, "c2"],
[1303, 1, "03"],
[1356, 92, "d036244663035ef7a5da58ce80b1182d0b74598fc9667c634eff0e13d42e272c63b1c1cf5a10b4c007d5c6cab9a57e3edf56678083cbe20a7ad9f3e59bbd1eb47b393a1824993fe19cb2913ba15fa45b87a9409a0fc3fb135f899422"],
[1547, 16, "245758832a316f67807d887140105ae4"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "start of an entry not the end of the one before",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "628c2e6abf142176591038e4b59d14cad3d25f"],
[790, 1, "28"],
[1357, 113, "dc71d99ae2ac5c235a5e4422e950acf34bb9636639aaaf2b7d3d0e82feaaeb54872756ae0daa54d1165983b4ddd459e8d022785621bd7ec8214dbf48af6025bcf87706102af7bdee93658a2346afb56171f0dd1b98789234fbd7ddb89d610957639f3227ba7f84c79b99b366db5a168992"],
[1547, 16, "63323ac9840e883e8b34330d265bd35d"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "end of the last file not C",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1370, 5, "654164032d"],
[1547, 16, "605b52ce964d3cc730559ec9d70517d4"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "a byte of a file changed",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[1416, 1, "ee"],
[1547, 16, "9a5f0b6854201aa920a20da65af2cf1b"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "path .. and a padding byte not zero",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[576, 19, "5c8c2e6abc8b5bd0ea02bce7e5cdeb0d462e06"],
[790, 1, "d2"],
[1360, 78, "99ad835637585d4c30b52863395955fef4df7192b4ce7156d23f20ce090f8b69cd399e474e1a5ff6953514ecedab2ac5a4e3f912b34435fbc24b4fc15e463c84ad3804ef94bc9f55c42cc03ea7cc"],
[1546, 17, "2708f4c808466d6fdf78eefba7fd30ec71"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEAD_INVALID",
"step": 17
},
{
"name": "path .. and the next STREAM chunk corrupt",
"spec": true,
"dkc": {
"base": "format3_tree.dkc",
"edits": [
[1358, 1, "2b"],
[1406, 26, "b2530de02226048b42042bfd246f7b537924806f283a8717576c"],
[66315, 16, "401e446ca7dd8e355a937ec755ba5502"],
[66431, 1, "e5"]
]
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:27Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "path .. and a cut right after its chunk",
"spec": true,
"dkc": {
"base": "format3_tree.dkc",
"edits": [
[1358, 1, "2b"],
[1406, 26, "b2530de02226048b42042bfd246f7b537924806f283a8717576c"],
[66315, 16, "401e446ca7dd8e355a937ec755ba5502"],
[66331, 20496, ""]
]
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:27Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_INTEGRITY",
"step": 17
},
{
"name": "security of version 2 opens with the verdict X",
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[812, 1, "1f"],
[1547, 16, "e7c6c4199c2d44d32f2ae14f7672fb98"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "X",
"seal": "X",
"lines": [
"No se han podido comprobar la firma ni el sello: trátala como no firmada y sin fecha probada."
]
}
},
{
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
"name": "a signature of alg 4294967295 opens with the verdict F1",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"spec": true,
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
"dkc": {
"base": "format3_single.dkc",
"edits": [
[786, 6, "256f1f0dc845"],
[813, 112, "20d7c80e99276f0d6767c0ffe2047e36a800f14138362cf668c59ca29a1a1f162ade4fb9b820db3648268570b6070dbcaa0cd7e8ed3fff0b3f50e46a5a5caa7482df632e59283d406d7be4db6165ac2d6a49110f439275e7680f0120893b727b82bb8619c4115411d05a7feb5f9e09e5"],
[1547, 16, "e68dbdf94498d76dd170fa3bfbbe344f"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F1",
"seal": "S0",
"lines": [
"No se ha comprobado ninguna firma: trátala como no firmada."
]
}
},
{
"name": "a signature of alg 1 that does not verify opens with the verdict F2",
Test data: security.json in the context of a capsule, with v0.12 verdicts The review found that security.json said spec 0.11 and still gave the verdicts of a reader of v0.10, evaluated without context, contrary to what section 76 announced (D1). - security.json carries the context of a capsule, its commitments and the time of its round, and each case its verdicts and the lines of the official SDK in it: a signature of alg 1 that does not verify is F2, a token of seal_type 2 that is not DER is S2, and new cases give a valid signature of alg 1 (F4) and alg and seal_type 4294967295 (F1, S1). - mutations.json: the signature of alg 1 that does not verify (F2) is a case of 64, and the seal that opens with S1 uses seal_type 4294967295, not seal_type 1, which a later version may define. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"spec": true,
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"dkc": {
"base": "format3_single.dkc",
"edits": [
[786, 6, "216f1f0dc845"],
[813, 108, "20d7cc0e993c91aab889d0b6d3047e36a800f14138362cf668c59ca29a1a1f162ade4fb9b820db364835cc21852777deaa0cd7e8ed3fff0b3f50e46a5a5caa7482df632e59283d406d7be4db6165ac2d6a49110f439275e7680f0120893b727b82bb8619c4115411d05a7feb"],
[1547, 16, "cf30e8d64e74222c7df80666b05eade7"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
"signature": "F2",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"seal": "S0",
"lines": [
Signature plan, steps 3 and 4: EncryptFiles signs and Open checks, alg 1 AreaLen is 32 KiB, and LargeArea asks for 64 KiB. EncryptOptions.AuthorKey signs inside sealer.write, through a prepare hook that gets the final control: SECURITY_CBOR and the frame are built and evaluated with the rules of the reader before anything is written. OpenOptions.AuthorKeys feeds EvaluateSecurityIn from openBody with control_commit, head_digest and the round time: F4, F3 with a saved key, F2 when it does not verify. The fixtures of v0.10 keep the area of 512 (AreaUnit). The two "unsupported" fixtures use alg 4294967295, since a random alg 1 is now F2. Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
6 days ago
"La firma no corresponde a este contenido."
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
}
},
{
Test data: security.json in the context of a capsule, with v0.12 verdicts The review found that security.json said spec 0.11 and still gave the verdicts of a reader of v0.10, evaluated without context, contrary to what section 76 announced (D1). - security.json carries the context of a capsule, its commitments and the time of its round, and each case its verdicts and the lines of the official SDK in it: a signature of alg 1 that does not verify is F2, a token of seal_type 2 that is not DER is S2, and new cases give a valid signature of alg 1 (F4) and alg and seal_type 4294967295 (F1, S1). - mutations.json: the signature of alg 1 that does not verify (F2) is a case of 64, and the seal that opens with S1 uses seal_type 4294967295, not seal_type 1, which a later version may define. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"name": "a seal of seal_type 4294967295 opens with the verdict S1",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"spec": true,
"dkc": {
"base": "format3_single.dkc",
"edits": [
Test data: security.json in the context of a capsule, with v0.12 verdicts The review found that security.json said spec 0.11 and still gave the verdicts of a reader of v0.10, evaluated without context, contrary to what section 76 announced (D1). - security.json carries the context of a capsule, its commitments and the time of its round, and each case its verdicts and the lines of the official SDK in it: a signature of alg 1 that does not verify is F2, a token of seal_type 2 that is not DER is S2, and new cases give a valid signature of alg 1 (F4) and alg and seal_type 4294967295 (F1, S1). - mutations.json: the signature of alg 1 that does not verify (F2) is a case of 64, and the seal that opens with S1 uses seal_type 4294967295, not seal_type 1, which a later version may define. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
[786, 6, "e06f1f0dc845"],
[813, 45, "21d78f0f99276f0d6767c0ffe2265c148a22d3631a140ed44ae7be80b8383d3408fc6d9b9a02f9146a04a75294"],
[1547, 16, "8731c709cc638cee980df229ab30e569"]
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F0",
"seal": "S1",
"lines": [
"Sin firma de autor.",
"Lleva un sello de tiempo que esta versión no sabe comprobar: aquí no prueba nada."
]
}
},
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
{
"name": "the signature of alg 1 altered opens with the verdict F2",
"spec": true,
"dkc": {
"base": "format3_signed.dkc",
"edits": [
[867, 1, "b2"],
[35595, 16, "c3fd532dd5a51df534f11289765b71cb"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F2",
"seal": "S0",
"lines": [
"La firma no corresponde a este contenido."
]
}
},
{
"name": "the signature of alg 1 removed opens with the verdict F0",
"spec": true,
"dkc": {
"base": "format3_signed.dkc",
"edits": [
[786, 6, "5f714ecf7f7d"],
[813, 108, "45e7568b60c4234cbfaa510a7a716891a089ede67b438fc5ee3a170425910e84ab50afcecd4905b01cce51ffc86f71a2562eec34a94226e2ec508c8276f021f0d8655e5e731d8c00ab12bd7b407fe32e2901a9a5818eaeaf9454d8fad6d4c0e6c5826a4dffc3d51914104c46"],
[35595, 16, "76c5c411252762afa3787ff884f80678"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F0",
"seal": "S0",
"lines": [
"Sin firma de autor."
]
}
},
{
"name": "the signature of alg 1 made again with another key opens with the verdict F4 of that key",
"spec": true,
"dkc": {
"base": "format3_signed.dkc",
"edits": [
[822, 98, "2d41715eb4c045f6fe99b11c7ce00dc0aa3b4bd986f4e2e3e93952e571d3dc2ccc09bf65b50c2d1018affcf73b6ab29ba74d5764ea45e9816ed9581db69c95731a233908e879ee91019df577aa7ced8c2b90fa49bcc77051d6e3a0318d539e41912e"],
[35595, 16, "2f07443d5eb70a88c96e9414bb757fee"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F4",
"seal": "S0",
"lines": [
"Firmado con la clave dkauthor1sug8kfx94r29vam52ann7m7g96gzcnluzlaxvj9ej6wzswxkdscq9nxk89. No prueba quién la tiene."
]
}
},
{
"name": "the signature of alg 1 transplanted to another capsule opens with the verdict F2",
"spec": true,
"dkc": {
"base": "format3_unsigned.dkc",
"edits": [
[786, 6, "8a7c6cd4ca3b"],
[813, 108, "6f815f8214c83aad66a057570616cd6f9aea8cb7ad32238a8a6297d53476c1728208f53c3ba99ad68fc013c307e5dd10b100a72938bd75e6c6b17154d6068b81d4aa569260d496a94f2693ad0053ee322bfa26ef03f28f940a3edd1596b0f26c1aec6cd3998c0797b5ab658c"],
[35595, 16, "60083fba2123edec5ad00eed5c170759"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F2",
"seal": "S0",
"lines": [
"La firma no corresponde a este contenido."
]
}
},
{
"name": "a key of 31 bytes in a signature of alg 1 opens with the verdict F1",
"spec": true,
"dkc": {
"base": "format3_signed.dkc",
"edits": [
[786, 1, "c8"],
[815, 7, "3e2860c52214a0"],
[853, 68, "1e96119046917c588cd677c3bbd7309d61dc480b7b1065eaefc733410e1693561c8efebf417782cf3e430ce69a89858edd59926dbb1d3fde98b44e6db86e99a082374746"],
[35595, 16, "f193a72da1dcb9b65be07b3717212273"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F1",
"seal": "S0",
"lines": [
"No se ha comprobado ninguna firma: trátala como no firmada."
]
}
},
{
"name": "a signature of 65 bytes of alg 1 opens with the verdict F1",
"spec": true,
"dkc": {
"base": "format3_signed.dkc",
"edits": [
[786, 1, "ca"],
[815, 1, "3c"],
[856, 1, "be"],
[35595, 16, "547f05ee603ecf3cbe3faff4289299dc"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F1",
"seal": "S0",
"lines": [
"No se ha comprobado ninguna firma: trátala como no firmada."
]
}
},
{
"name": "the area widened to 64 KiB after signing opens with the verdict F4 and the same AUTHOR_MESSAGE",
"spec": true,
"dkc": {
"base": "format3_signed.dkc",
"edits": [
[574, 21, "691bbfdaa29b819a0f6e2a898089b5095f781eeb73"],
[779, 34832, "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
[35611, 0, "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
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ok",
"step": 0,
"verdicts": {
"signature": "F4",
"seal": "S0",
"lines": [
"Firmado con la clave dkauthor1jtf6s2c7ywr6scx40jhye32sj82rjp8a6cj7f54jsh36rd96vaqqcg54xg. No prueba quién la tiene."
]
}
},
{
"name": "the public note changed in PUBLIC_HEADER",
"spec": true,
"dkc": {
"base": "format3_note.dkc",
"edits": [
[184, 1, "6e"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Test data: the mutations of a signature of alg 1 and of the public note The review found that the cases that spec 64 lists for the signature of alg 1, the area and the public note were tested in Go and not exported to testdata, where a second implementation reads them (D3). mutations.json adds eight, from format3_signed, format3_unsigned and a new format3_note: - the signature altered (F2), removed (F0), made again with another key (F4 of that key) and transplanted to another capsule (F2); - a key of 31 bytes and a signature of 65 (F1); - the area widened to 64 KiB after signing (F4, the same AUTHOR_MESSAGE); - the public note changed in PUBLIC_HEADER (ERR_HEADER_BINDING, step 15). The frozen cases of the corpus do not change. The records of the fixtures give the extensions of the header as it is written, the note included. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
6 days ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_HEADER_BINDING",
"step": 15
},
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
{
"name": "format 3 time_only relabeled format 1",
"spec": false,
"dkc": {
"base": "format3_single.dkc",
"edits": [
[4, 1, "01"]
]
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
{
"name": "format 3 time_and_key relabeled format 2, with the identity",
"spec": false,
"dkc": {
"base": "format3_time_and_key_portable.dkc",
"edits": [
[4, 1, "02"]
]
},
"identities": [
"AGE-SECRET-KEY-1GTJSK6HRKTCQK70SSCYL0LS3LWSJ2MZ5CMA2PJ0VJFVTTAH5CPVSYGJ20R"
],
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_UNSUPPORTED_VERSION",
"step": 14
},
{
"name": "format 3 time_and_key relabeled format 2, with the .dkk",
"spec": false,
"dkc": {
"base": "format3_time_and_key_portable.dkc",
"edits": [
[4, 1, "02"]
]
},
"dkk": "444b4b31010000000000008ca70073646174656b6579732d6163636573732d6b657901010250bdb483fba42daf0b409f44d23033f362035029737cb54ad5310734ecf20720c2d317046678323535313905582042e50b6ae3b2f00b79f08609f7fe11fba1256c54c6faa0c9ec9258b5f6f4c05906a1005820680d29962e575689a31543df28433dae7737abd9a793e9cae92ef40920d09636",
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
"source": "supplied",
Format 3, step 6d: the mutations of format 3 The mutation corpus of spec 64 gains format 3: 209 cases, 169 of the spec. - The 33 mutations of the first two lists on format3_single and format3_time_and_key_portable, named "format 3: ...", with a sibling written by EncryptFiles and built capsules that hold a BODY. - The list of format 3, 47 cases: one for each value of a line with several (AREA_LEN 0, 511, 513 and 66048; SECURITY_LEN 0 and 513; HEAD_LEN 0 and 2^24 + 1), and the three that open without a code, with the verdicts X, F1 and S1. VERSION 4 is "format 3: version changed", as in format 2. Each seals BODY again with FK_PAYLOAD and the nonce of its fixture, and the control with the new L when L changes; the two that need a head followed by another STREAM chunk derive from format3_tree, whose comment takes the bytes the path loses so that only the head and its chunk change. - Further cases: format 3 relabeled 1, and time_and_key relabeled 2 with the identity and with the .dkk. - A mutation may expect the capsule to open with its verdicts; the exported case records them, with the result ok at step 0. - Splice gives an edit for each run of changed bytes, runs closer than 16 bytes merged, and one more for what one side has beyond the other: a head sealed again changes its bytes and the tag of its chunk, 64 KiB apart. Earlier cases are written with more edits and give the same capsules. The corpus is 706 KB, 476 KB of them the capsule of 65536 implicit folders, whose head is 235 KB. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 week ago
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_ACCESS_INVALID",
"step": 9
Release object, release in hand and step 9.c option B (spec v0.15 draft) The release of a round becomes a file, .dkr: a release object in deterministic CBOR, {0: "datekeys-release", 1: 1, 2: chain_hash, 3: round, 4: signature}, which provider.EncodeRelease writes and DecodeRelease reads with its layers (size, type and version, schema). provider.ParseRelease also reads drand's JSON as the input of the caller. Verify checks the chain hash a release names before its round and its signature, with ERR_PROFILE_MISMATCH. provider.Archive reads a local release archive, the informative format of the draft. capsule.OpenOptions.Release takes a release in hand, a provider.Supplier, exclusive with Source: Open does not compare it with the clock (step 9.c, option B) and reports a clock behind it in Opened.ClockBehind; a network source is still never asked before the round time. The CLI gains decrypt -release FILE (.dkr, drand's JSON or a local archive), decrypt -save-release FILE.dkr and the command release, which fetches, verifies and saves the .dkr without opening the capsule. Test data: vectors/release.json, releases/<round>.dkr for rounds 1000, 1001, 1004 and 2000, and a local archive of rounds 1000 to 1004. In mutations.json every case says its source, "supplied" or "network"; the case "round not reached yet", a release in hand, now opens, and four cases are added: the same with a network source, a release of another round from a network source, and two release objects of another chain. SpecVersion stays 0.14 until the author approves the draft. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1 day ago
},
{
"name": "round not reached yet, from a network source",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39"
},
"source": "network",
"now": "2023-08-23T15:59:23.999999999Z",
"registry": "default",
"network": false,
"frozen": false,
"error": "ERR_RELEASE_UNAVAILABLE",
"step": 9
},
{
"name": "release of another round, from a network source",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41"
},
"source": "network",
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_RELEASE_UNAVAILABLE",
"step": 9
},
{
"name": "release object of another chain",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1000,
"signature": "b44679b9a59af2ec876b1a6b1ad52ea9b1615fc3982b19576350f93447cb1125e342b73a8dd2bacbe47e4b6b63ed5e39",
"chain_hash": "52db9ba70e0cc0f6eaf7803dd07447a1f5477735fd3f661792ba94600c84e970"
},
"source": "supplied",
"now": "2023-08-23T15:59:24Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_PROFILE_MISMATCH",
"step": 10
},
{
"name": "release object of another chain and another round, with a clock behind",
"spec": false,
"dkc": {
"base": "time_only.dkc",
"edits": []
},
"release": {
"round": 1001,
"signature": "b33bf3667cbd5a82de3a24b4e0e9fe5513cc1a0e840368c6e31f5fcfa79bea03f73896b25883abf2853d10337fb8fa41",
"chain_hash": "d2db9ba70e0cc0f6eaf7803dd07447a1f5477735fd3f661792ba94600c84e971"
},
"source": "supplied",
"now": "2023-08-23T15:59:23.999999999Z",
"registry": "default",
"network": true,
"frozen": false,
"error": "ERR_PROFILE_MISMATCH",
"step": 10
Hand-written CBOR codec and shared test vectors (plan steps 2b and 3) Step 2b: package codec is rewritten without reflection or struct tags. A strict Decoder accepts only the spec §58 profile, Unmarshal decodes, re-encodes and compares, Peek reads the type tag and version, and Walk is a bounded iterative helper for vectors and fuzzing. Every schema has its own hand-written encoder and decoder that checks all CDDL rules before the fields with their own error codes. github.com/fxamacker/cbor/v2 and github.com/x448/float16 are gone; nothing replaces them. Valid objects encode and decode exactly as before (1.34 million differential verdicts); the invalid-input differences are documented in CHANGELOG and traceability decision 12. A review found and fixed an access_policy check that truncated to uint8. Step 3: testdata gains vectors/cbor.json (generic and per-schema CBOR vectors), vectors/mutations.json (the 55-case mutation corpus, replayable offline), vectors/inspect_differential.json (1,825 fixed-seed mutations with the Go verdict) and one inspect -json golden per fixture, all regenerated byte-identically by genfixtures and documented in testdata/README.md for second implementations. Gate green with 90 s of fuzzing per target on all 15 targets; codec at 100 % coverage; pre-existing testdata byte-identical. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2 weeks ago
}
]
}

Powered by TurnKey Linux.