Found while wiring the measurement domain (quoin#478, PR #497), pre-existing and not caused by it — it fails identically against a quoin-core built from a branch without that change, and passes when run alone.
tests/core-exec-e2e.test.ts > accepts the real binary under its own pinned digest exceeds its 5 s timeout when the whole suite is running. The test sha256s a debug quoin-core binary of roughly 400 MB and then spawns it twice; under parallel load that does not fit in 5 s.
It is a test-harness bound, not a product defect: the digest pin itself is correct and the binary is the right one. Two honest options, in preference order:
- Point the test at a release binary, so the thing being hashed is the size the pin is actually for.
- Raise this one test's timeout with a comment saying what it is waiting on — a fixed cost that scales with binary size, not a race.
Do not delete or skip it. QUOIN_EXPECTED_CORE_SHA256 is the bytes-pinning guard that src/quire/exec.ts learned the hard way (quoin#164 and its siblings), and this is the only test that exercises it end to end.
Parent: #373
Found while wiring the measurement domain (quoin#478, PR #497), pre-existing and not caused by it — it fails identically against a
quoin-corebuilt from a branch without that change, and passes when run alone.tests/core-exec-e2e.test.ts > accepts the real binary under its own pinned digestexceeds its 5 s timeout when the whole suite is running. The test sha256s a debugquoin-corebinary of roughly 400 MB and then spawns it twice; under parallel load that does not fit in 5 s.It is a test-harness bound, not a product defect: the digest pin itself is correct and the binary is the right one. Two honest options, in preference order:
Do not delete or skip it.
QUOIN_EXPECTED_CORE_SHA256is the bytes-pinning guard thatsrc/quire/exec.tslearned the hard way (quoin#164 and its siblings), and this is the only test that exercises it end to end.Parent: #373