Rongsong Shen
d26ff8870d
The buglog was one 27K-line JSON file whose fields covered only symptom,
root cause and fix, as free text: 2% of entries named the code version a
bug was found on, 12% the build profile, 36% a way to reproduce, and 58% of
the file was OpenWolf hook auto-detections ("Missing guard clause",
"Function not marked async" on .md files).
- .claude/skills/buglog/bug-record.md: the record standard. Eight
sections (Summary as the title, Description, Environment, Reproduction;
Root cause, Fix, Verification, References once fixed), each with the
content it must contain -- e.g. Environment needs the OS, the libc the
failing binary loads (from its ELF interpreter: glibc 2.42 in the nix
shell vs 2.40 on the host), toolchain, build profile and commit.
- .wolf/bugs/<id>.md: 500 records migrated from buglog.json; every field
verified to round-trip. They carry legacy: true and an explicit "not
recorded" where the old log had no such field; component derived from
file paths, severity from keywords (else unknown), fix_commits from
hashes in status_note/fix that git resolves (91 of 463 fixed).
- .wolf/buglog-archive.json: the 710 hook auto-detections, verbatim.
- .wolf/buglog.json is now an index generated from the records, keeping
the fields OpenWolf's hooks and `openwolf bug search` read.
- scripts/buglog.py rewritten: new (template + Environment collected from
the machine and --binary), check, status (refuses a status the record
doesn't support; fixed needs Root cause, Fix, Verification and
--commit hashes git resolves), note, recur, set, show, list, triage,
env, index. 29 unittest cases in scripts/tests/test_buglog.py.
- scripts/git-hooks/pre-commit runs `check --staged`: incomplete staged
records, a legacy record hand-changed to fixed without the fix-time
sections, a new record claiming legacy: true, or a stale index refuse
the commit.
- scripts/gitea-sync.py reads the records: issue body = the record's
sections plus linked fix commits (origin remote, GITEA_COMMIT_REPO to
override), and a body hash so an issue's text is refreshed when the
record changes.
- bug-1866 rewritten to the full standard as the worked example.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
3.4 KiB
| id | status | status_updated | status_note | component | severity | timestamp | last_seen | occurrences | files | tags | related | fix_commits | legacy | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| bug-1603 | fixed | 2026-09-24 | build | unknown | 2026-08-20T09:20:00Z | 2026-09-19T09:00:00Z | 1 |
|
|
|
true |
bug-1603: build-time-versioning task 2.2 REJECTED by reviewer: nix build .#xcecore_debug (no flags, the…
Description
build-time-versioning task 2.2 REJECTED by reviewer: nix build .#xcecore_debug (no flags, the exact documented/task-Observable command) failed at evaluation with error: attribute 'currentTime' missing; also the raw epoch-seconds string didn't satisfy Requirement 1.2's absolute-UTC-form requirement.
Environment
Not recorded (migrated from buglog.json).
Reproduction
Not recorded (migrated from buglog.json).
Root cause
First attempt computed myBuildTime = builtins.toString builtins.currentTime at flake-evaluation time (src/flake.nix), threaded through nixSupport/default.nix into xce.nix/xcecli.nix's mesonFlags. builtins.currentTime is stripped entirely under a flake's default pure evaluation (not just fixed to a dummy value -- the attribute itself is removed from builtins), so any nix build/eval that forces it fails unless --impure is passed. This also produced a raw epoch-seconds string, not an ISO-8601 UTC string like xce-build.sh's (task 2.1) date -u +%Y-%m-%dT%H:%M:%SZ output.
Fix
Moved the build-time computation OUT of Nix expression evaluation and INTO each derivation's build phase. xce.nix and xcecli.nix each gained a preConfigure hook that runs date -u +%Y-%m-%dT%H:%M:%SZ (stdenv's bundled coreutils, always available in the build sandbox, unaffected by pure-eval, no --impure/git/network needed) and appends -Dbuild_time=$buildTime to mesonFlagsArray (verified against the real nixpkgs meson setup-hook.sh: mesonConfigurePhase runs runHook preConfigure BEFORE it reads mesonFlags/mesonFlagsArray via concatTo, so the appended flag is picked up at configure time). The hook exit 1s if date produces empty output, mirroring xce-build.sh's own non-empty-check discipline. Since build_time is no longer an eval-time concept, myBuildTime was removed entirely from flake.nix/nixSupport/default.nix/the xce.nix+xcecli.nix function signatures and their eval-time asserts -- only myCommitId (a genuine eval-time/flake-level value from self.shortRev/self.dirtyShortRev) still threads through. Live-verified: nix eval --json .#packages.x86_64-linux.xcecore_debug.mesonFlags (no --impure) succeeds; nix build .#xcecore_debug (no flags) proceeds through evaluation+configure+570/614 ninja targets (all major C binaries linked) before failing only on an unrelated pre-existing sandbox network restriction (autoscaler/bcctool's Go module fetch from goproxy.cn, nothing to do with this fix); build log shows -Dbuild_time=2026-08-20T09:13:29Z reaching meson configure and meson's own summary printing build_time : 2026-08-20T09:13:29Z (real, non-empty, ISO-8601 UTC). Same confirmed for xcecli.nix (-Dbuild_time=2026-08-20T09:15:46Z). commit_id=bca6397-dirty confirmed unaffected/still working in both.
Verification
Not recorded (migrated from buglog.json).
References
Not recorded (migrated from buglog.json).