ci(tests): run Windows tests on Python 3.10 and 3.13 - #4473
Closed
feiiiiii5 wants to merge 1 commit into
Closed
Conversation
Pin tests-windows to 3.13 only exercises the one Windows Python where the clock is fine-grained, leaving coarse-clock bugs (like openai#4392) invisible for the other supported versions. Matrix the job across 3.10 (the declared minimum, coarse clock) and 3.13, matching the shape proposed in openai#4472 at 2x the current Windows cost. Signed-off-by: fei <204683769+feiiiiii5@users.noreply.github.com>
Member
|
Thanks for sending this. #4475 resolves the issue. |
Author
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
tests-windowspins a single Python version (3.13), which is exactly the version where the Windows clock became fine-grained — so the whole class of coarse-clock bug (e.g. #4392, 37 tests failing on a clean checkout while CI stayed green) is invisible on Windows for the other supported versions. This matrices the job across"3.10"(the declared minimum, on the coarse-clock side) and"3.13"(what runs today), using the shape suggested in the issue: 2x the current Windows runner cost instead of 5x.fail-fast: falsekeeps both legs reporting, and thedetect-changesgate is preserved unchanged.packaged-contract-windowsis intentionally untouched (packaging smoke, not test coverage).Test plan
Workflow-only change; verified locally:
tests.ymlparses as valid YAML and thetests-windowsjob carriesstrategy.matrix.python-version == ["3.10", "3.13"]withfail-fast: false, andSetup uvconsumes${{ matrix.python-version }}(asserted programmatically).testsjob, and no other job references the changed keys.Final proof is CI itself: this PR's checks page should show two
tests-windowslegs (3.10 and 3.13), both honoring thedetect-changesgate.Issue number
Fixes #4472
Checks
.agents/skills/code-change-verification/scripts/run.sh/reviewbefore submitting this PRThe verification-stack checkboxes are left unchecked honestly: this diff touches only a CI workflow file, and the applicable verification is the YAML/structure check described in the test plan plus the workflow legs on this PR.