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>
2.8 KiB
| id | status | status_updated | status_note | component | severity | timestamp | last_seen | occurrences | files | tags | related | legacy | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| bug-1595 | fixed | 2026-09-24 | scheduler | unknown | 2026-08-19T08:00:00Z | 2026-09-19T09:00:00Z | 1 |
|
|
true |
bug-1595: User-reported: in each step, it would be better to show the decisions/events that occur in the…
Description
User-reported: in each step, it would be better to show the decisions/events that occur in the current session in the sequence they actually occur -- currently only decisions are printed, input events (JOB_PUSH, JOB_STATE_CHANGE, WORKERS_PUSH, etc.) are completely silent unless a breakpoint happens to match one.
Environment
Not recorded (migrated from buglog.json).
Reproduction
Not recorded (migrated from buglog.json).
Root cause
sim_loop.c already calls xce_repl_check_event()/xce_repl_on_event() for every decoded ReplayEvent, in stream order, strictly before it joins the batch xce_core_push_events() later applies -- but xce_repl_on_event() (repl_break.c) only ever prints anything when a breakpoint matches; there is no unconditional per-event echo. sim_print_decision() (sim_loop.c), by contrast, IS called unconditionally for every drained decision. Net effect: an operator using step/continue without breakpoints sees decisions appear with no visible cause -- the events that produced them (a job arriving, a job's state changing, a worker reporting in) are invisible.
Fix
Fixed as part of task 12.5. xce_repl_on_event() (repl_break.c) now unconditionally prints a one-line human-readable summary of every event before running the (unchanged) breakpoint-match check -- e.g. "[sim] event: JOB_PUSH njobs=2 jobids=[<fmt_jobid>, <fmt_jobid>]", "[sim] event: JOB_STATE_CHANGE jobid=<fmt_jobid> state=RUN", "[sim] event: WORKERS_PUSH nworkers=1 names=[worker-1]", one line per xce_core_event_type_t kind. Since xce_repl_on_event() is only ever invoked when sim->interactive (via repl.h's existing gate), this adds zero output to batch --sim mode. repl_job_state_name() (repl_info.c) was made non-static and declared in repl_internal.h so both repl_info.c's info job and repl_break.c's new event summary render the same state name for JOB_STATE_CHANGE. Printed order is now event(s) (as decoded) -> BEGIN_ROUND -> decision(s) (as drained), matching sim_loop.c's actual processing order. New tests in test_repl_breakpoints_inspection.c step through a fixture with a JOBS_PUSH and a JOB_STATE_CHANGE event and assert both event summary lines appear, in order, before the resulting decision line.
Verification
Not recorded (migrated from buglog.json).
References
Not recorded (migrated from buglog.json).