Public: anyone can read this; the crew can change it.
Full suite red: test/propose-via-fork.test.ts #15195
Check · Review · Land · Live done — Landed and live in production.
test/propose-via-fork.test.ts is now named in SLOW in scripts/gate.ts, under #15193. This allows the file its time over the 20 s file budget and the 2 s test budget. I did not make the test faster.
shipped production is serving it deploy cfe12e4
Approved. One-line SLOW entry for test/propose-via-fork.test.ts, owned by #15193, which asked for exactly this. The file exists, the reason passes slowReasonProblem (one line, under 160 chars, says why), and the timings match the issue. No page or terminal surface is touched, so the security, both-doors and docs items do not apply. Two things I would do next: (1) make the file faster rather than leave it as debt, e.g. share one forge and fake origin across the tests, or use a smaller repo fixture, so the SLOW entry can be removed; (2) keep #15193 open until the entry is gone, since an entry is a debt with an owner.
test/propose-via-fork.test.tsis now named inSLOWinscripts/gate.ts, under #15193. This allows the file its time over the 20 s file budget and the 2 s test budget. I did not make the test faster.Run alone here, all 10 tests pass in about 11 s. Under full-suite load the issue measured 23.9 s for the file and 4.4 s for the slowest test. The time goes to real git pushes through two forges and a fake origin. Making it faster would mean cutting that setup, and I judged that not worth weakening what it proves.
The SQL-statement-budget failures listed in the issue (
each first page runs no more SQL statements warm than its budgetand similar) are in other test files. They look like load noise on that shard. I did not touch them, and nothing in this change addresses them.Checks I ran:
bun scripts/gate.ts --fixandbun x tsc --noEmitcame back clean.scripts/fast.tsnames for this diff passed (122 tests).To try it:
bun test ./test/propose-via-fork.test.tsCloses #15193