ci: run the pre-push hooks sequentially
CI / check (push) Skipped
CI / e2e (push) Skipped
CI / check (pull_request) Successful in 2m30s
CI / e2e (pull_request) Successful in 6m23s

`go test -race ./...` saturates every core for ~47s, and the UI tier
it was sharing them with is a real Chromium with wall-clock timeouts.
So the browser lost, at random: setup took 106s inside the hook
against 63s standalone, and a different suite failed on each run --
three failing to fetch setup.ts from Vitest's own dev server once, a
15s "did not mount itself" the next time -- against a suite that
passes 898/898 five times running on its own.

That reads as "your branch broke the frontend" when nothing is wrong,
which is the most expensive kind of false negative: the next person
bisects a change that was never at fault. It cost two pushes here
before the summary line gave it away.

Sequential costs about 15s.

Closes #128
This commit is contained in:
2026-08-19 09:16:43 -04:00
parent 2b84bc53e9
commit 61d549a9d5
+10 -1
View File
@@ -63,8 +63,17 @@ pre-commit:
root: "frontend/"
run: node scripts/check-css-literals.mjs
# Deliberately sequential, unlike pre-commit. `go test -race`
# saturates every core for the better part of a minute and the UI tier
# is a real browser with wall-clock timeouts, so run together the
# browser loses: setup took 106s inside the hook against 63s
# standalone, and a different suite failed each time -- three suites
# failing to fetch setup.ts from Vitest's own dev server on one run, a
# 15s "did not mount itself" on the next, against a suite that passes
# 898/898 on its own. A gate that fails at random is not a gate. The
# ~15s saved is not worth it.
pre-push:
parallel: true
parallel: false
commands:
go-test:
glob: "*.go"