Skip to content

1.5 release #100

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Mar 8, 2024
Merged

1.5 release #100

merged 6 commits into from
Mar 8, 2024

Conversation

asmeurer
Copy link
Member

@asmeurer asmeurer commented Mar 8, 2024

No description provided.

@asmeurer asmeurer enabled auto-merge March 8, 2024 05:35
@asmeurer asmeurer temporarily deployed to publish-package March 8, 2024 05:37 — with GitHub Actions Inactive
@rgommers
Copy link
Member

rgommers commented Mar 8, 2024

I'm a little confused why this PR is still open but the 1.5 release tag already up on PyPI?

@asmeurer
Copy link
Member Author

asmeurer commented Mar 8, 2024

Because a (flaky) test failed on the tag, which is the same commit as the PR, so it didn't merge.

@rgommers
Copy link
Member

rgommers commented Mar 8, 2024

Hmm, that seems a bit unhealthy. Don't rely on auto-merge for future tags I guess, but only push them when the release commit to be tagged is merged?

@asmeurer asmeurer merged commit d10bf57 into data-apis:main Mar 8, 2024
@asmeurer
Copy link
Member Author

asmeurer commented Mar 8, 2024

I do it that way on purpose, because if the publish build fails I have to delete the tag and try again. That happened with the last release because I forgot to bump the version, and it almost happened again this release. I prefer for "PR merged" to mean "the release is complete". If I see the PR is open that means there is still something left to do. I don't "rely" on auto-merge, but rather just use it as a way of saying "everything is good from my end, if the tests pass you can merge".

It's annoying that GitHub stops a PR merge because of tag builds, since they are on the same commit. Of course the real issue here is the flakeyness of the tests, but that is just the nature of this project unfortunately.

@rgommers
Copy link
Member

rgommers commented Mar 9, 2024

Makes sense, thanks for the context. Pushing the tag before everything else is 100% complete never works if there's builds that can fail. The risk here seems rather low that the tag has to be moved.

Making the publish job run only after the other jobs all complete would be an alternative - but that slows down regular CI, so also not ideal.

@asmeurer
Copy link
Member Author

I think the publish build doesn't actually get cancelled by other builds, but it's hard to tell.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants