• Joined on 2026-03-08
ltms closed pull request fleet/fleetd#287 2026-09-04 06:14:11 +02:00
fleetd #284: free backend-error capacity
ltms closed issue fleet/fleetd#285 2026-09-04 06:14:08 +02:00
seedTrustDialog writes the trust seed to fleetd's own home when memberHerdrSocket is set, so the member never sees it
ltms closed issue fleet/fleetd#284 2026-09-04 06:14:05 +02:00
A BACKEND_ERROR member holds its seat forever: counted as live, never reaped, never reclaimable
ltms commented on issue fleet/fleetd#285 2026-09-04 06:13:53 +02:00
seedTrustDialog writes the trust seed to fleetd's own home when memberHerdrSocket is set, so the member never sees it

Merged as 75b1508, corrected in 94f50e5. Real merge built green at 1296 tests.

The defect is real and reachable in code, as the worker reported. seedTrustDialog was static, so it…

ltms commented on issue fleet/fleetd#284 2026-09-04 06:13:34 +02:00
A BACKEND_ERROR member holds its seat forever: counted as live, never reaped, never reclaimable

Merged as dab9645, corrected in 94f50e5. Real merge built green at 1296 tests.

Option 2 was chosen (keep the roster entry, stop it counting), and the worker extended it from `BACKEND_ERROR…

ltms pushed to main at fleet/fleetd 2026-09-04 06:13:11 +02:00
94f50e507a #284/#285: one reclaimable rule, one group-share helper
75b15086b0 Merge #285: the workspace-trust seed refuses rather than write fleetd's own home
dab9645906 Merge #284: a BACKEND_ERROR or FAILED member no longer holds a spawn seat
7e49c6eca2 #285: seedTrustDialog refuses under memberHerdrSocket instead of seeding fleetd's own home
11cbfa79b4 fleetd #284: free failed-session capacity
Compare 6 commits »
ltms commented on pull request fleet/fleetd#291 2026-09-04 06:12:46 +02:00
#285: seedTrustDialog refuses under memberHerdrSocket instead of seeding fleetd's own home

Merged locally. One change on top of the branch: the new shareTrustJsonWithGroup helper moved into EnvAllowListScrub as shareFileWithGroup(Path, String). Reasoning is on #285.

ltms commented on pull request fleet/fleetd#287 2026-09-04 06:12:43 +02:00
fleetd #284: free backend-error capacity

Merged locally. Posting this here because the worker had no gh and could not update its own PR body — the round-two commit 11cbfa7 extends the fix from BACKEND_ERROR to FAILED as well,…

ltms closed pull request fleet/fleetd#288 2026-09-04 06:00:35 +02:00
#283: fix two teardown-cleanup leaks in SessionManager
ltms closed issue fleet/fleetd#283 2026-09-04 06:00:32 +02:00
Two teardown leaks in SessionManager: an unguarded worktree removal, and a branch the spawn-failure catch forgets
ltms commented on issue fleet/fleetd#283 2026-09-04 06:00:29 +02:00
Two teardown leaks in SessionManager: an unguarded worktree removal, and a branch the spawn-failure catch forgets

Merged as c2c2746. Real merge built green at 1287 tests. I ran my own mutations rather than trusting the worker's:

Defect 1 — put release()'s worktree removal back to a bare…

ltms closed pull request fleet/fleetd#289 2026-09-04 06:00:07 +02:00
#282: fix a chained fleet_ask killing its own async ticket
ltms closed issue fleet/fleetd#282 2026-09-04 06:00:04 +02:00
A second fleet_ask in the same resumed turn kills its own async ticket: answer() completes it with QUESTION
ltms commented on issue fleet/fleetd#282 2026-09-04 06:00:01 +02:00
A second fleet_ask in the same resumed turn kills its own async ticket: answer() completes it with QUESTION

Merged as bfabe13, with a follow-up commit e93b5f6 correcting one claim. Full build of the real merge: 1288 tests, BUILD SUCCESS.

One claim in the PR does not hold

The PR says each…

ltms pushed to main at fleet/fleetd 2026-09-04 05:59:38 +02:00
e93b5f6512 #282: record that answer()'s QUESTION guard is defence in depth, not load-bearing
bfabe13e8f Merge #282 (PR #289): a chained second fleet_ask no longer kills its own async ticket
6ed70700a0 #282: don't let a chained fleet_ask kill its own async ticket
Compare 3 commits »
ltms pushed to main at fleet/fleetd 2026-09-04 05:56:32 +02:00
c2c2746922 Merge #283 (PR #288): guard release()'s worktree removal, delete the orphaned branch on spawn failure
19cacf5b62 #283: guard release()'s bare worktree remove; delete orphaned branch on spawn failure
Compare 2 commits »
ltms opened issue fleet/fleetd#290 2026-09-04 05:55:08 +02:00
Restore a test for reapIdle's per-session guard, which #283 left uncovered
ltms closed pull request fleet/fleetd#286 2026-09-04 05:53:07 +02:00
#281: pin handler authorization actions
ltms closed issue fleet/fleetd#281 2026-09-04 05:53:04 +02:00
Pin the Authz.Action every handler chooses, so the next #272 fails a test instead of shipping
ltms commented on issue fleet/fleetd#281 2026-09-04 05:53:01 +02:00
Pin the Authz.Action every handler chooses, so the next #272 fails a test instead of shipping

Merged as 0087645. Both properties hold, and I proved each one myself against the real merge rather than on the worker's word — deliberately using different mutations than the ones it…