Block a user
fleetd #513: fix TIMED_OUT_QUEUED javadoc — cancelled, not queued
MessageService's TIMED_OUT_QUEUED javadoc says the message is still queued; the code 12 lines of behaviour later cancels it
TurnListener implementations are the only thing maintaining the Injector's own invariant
Criterion 5, step 2: "the test passes" is a SUCCESS outcome, not an exception clause
Raised by the fleet01 lead, correcting the shape of my own wording in comment 16974. This is a…
MessageService's TIMED_OUT_QUEUED javadoc says the message is still queued; the code 12 lines of behaviour later cancels it
Rework needed on PR #564 — the new javadoc replaces a wrong claim with a narrower wrong claim
I verified the build on a tree merged with origin/main. That part is clean:
- merge of…
A lead cannot ack its own held peer mail, so every lead-to-lead message it polls is delivered a second time and costs a whole lead turn
TurnListener implementations are the only thing maintaining the Injector's own invariant
Sharpening acceptance criterion 5 — "red today" is not executable after the merge
Raised by the fleet01 lead, correcting wording of mine. This is a correction to the brief; it binds,…
LoopWatchdog.health() is public and nothing reads it — surface it without breaking the two scripts that parse /healthz
TurnListener implementations are the only thing maintaining the Injector's own invariant
Unblocked, plus one addition to Acceptance
Unblocked. #553 merged as f606fcc and is on main. The "merge that first" line in Out of scope is satisfied.
Addition to Acceptance —…
Fleetd's TurnListener.onDelivered ordering is load-bearing and untested: swapping two lines makes a throw strand the caller for its full timeout
This ticket's test is a regression guard, not a fix — and it can entrench the real defect
Raised by the fleet01 lead, whose finding this ticket already credits. Recording it here because…
A dead StatusPoller or SessionReaper can now be restarted, but nothing restarts it and nothing notices
A dead StatusPoller or SessionReaper can now be restarted, but nothing restarts it and nothing notices
Merged as part of #559. Verified by the lead, on a merged tree — not on the branch.
Setup
main at 7a3b2bb + branch bfac141, merged in a throwaway worktree = 5de807f.
Baseline:…
CLAUDE.md: the ticket is the pull channel, and the brief is write-once
ltms
deleted branch worker/544-loop-progress-watchdog-9f60b3-3 from fleet/fleetd
2026-09-12 11:11:48 +02:00
fleetd #544: progress watchdog for StatusPoller and SessionReaper loops
Fleetd's TurnListener.onDelivered ordering is load-bearing and untested: swapping two lines makes a throw strand the caller for its full timeout
7a3b2bb7ee
Merge pull request 'fleetd #552: warn instead of aborting when the post-restart mktemp fails' (#560) from worker/552-post-restart-mktemp-abort-bc2672-4 into main