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>
3.3 KiB
| id | status | status_updated | status_note | component | severity | timestamp | last_seen | occurrences | files | tags | related | legacy | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| bug-1724 | fixed | 2026-09-24 | xceadmin | unknown | 2026-08-28T00:00:00Z | 2026-09-19T09:00:00Z | 1 |
|
|
true |
bug-1724: 3 xceadmin live E2E tests…
Description
3 xceadmin live E2E tests (test_e2e_live_push.py::test_live_config_push_against_real_single_node_controller, test_e2e_live_ha_push.py::test_live_config_push_against_primary_reaches_success_path, test_e2e_live_ha_push.py::test_live_config_push_against_standby_returns_standby_rejection_not_redirect) always SKIPPED under the meson-driven build path (/xce-build xceadmin, nix build .#xcecore) with 'no xce_controller binary found under xce.build/build.release/build.dbg', regardless of whether a real xce_controller/xceadmin binary actually existed anywhere in the checkout.
Environment
Not recorded (migrated from buglog.json).
Reproduction
Not recorded (migrated from buglog.json).
Root cause
tests/_live_e2e_common.py's CONTROLLER_CANDIDATES/XCEADMIN_CANDIDATES/DEPLOY_HA_SH resolve paths relative to the test file's own on-disk location (Path(file).resolve().parents[2]). That works when pytest runs directly from the real checkout, but meson-build.sh copies the whole xceadmin test tree into an isolated out-of-tree repo-mirror before invoking pytest (to keep the meson build pure), and only ever symlinked ha/, src/configs, src/schema into that mirror -- never the sibling xce.build/build.release/build.dbg controller-binary directories, nor src/tools/ (deploy-ha.sh). So the mirror's own SRC_DIR (repo-mirror/src/) never had a controller subdir, a deploy-ha.sh, or a matching xceadmin binary at the expected relative path, no matter what was actually built in the real repo -- the skip was structural, not a signal about the real build state.
Fix
Two changes to meson-build.sh: (1) symlink each profile's controller subdirectory (xce.build/controller, build.release/controller, build.dbg/controller -- only the controller subdir, never the whole profile-root, since the profile-root can be an ancestor of the mirror itself when xceadmin is built in that same profile, which would create a self-referential symlink loop) into the mirror so CONTROLLER_CANDIDATES can find a real xce_controller wherever it was built; (2) symlink src/tools (checked-in, not a build-output dir, so a plain safe symlink like configs/schema) so DEPLOY_HA_SH resolves. For xceadmin's OWN binary, which genuinely is self-referential (the mirror lives inside that very build's own xceadmin output), routed around it via the existing XCEADMIN_BIN env-var override (resolve_binary_candidates in _live_e2e_common.py), set to $VENV_DIR/bin/xceadmin -- the binary this very meson-build.sh invocation just pip-installed -- before invoking pytest. Verified end-to-end via a real xce-build.sh xceadmin debug build.dbg --incremental run: all 3 tests now execute for real against a live controller/cluster instead of skipping.
Verification
Not recorded (migrated from buglog.json).
References
Not recorded (migrated from buglog.json).