Skip to content
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

Fixing CI on stable branch #88

Merged
merged 9 commits into from
Dec 12, 2024
Merged

Fixing CI on stable branch #88

merged 9 commits into from
Dec 12, 2024

Conversation

YannickJadoul
Copy link
Collaborator

@YannickJadoul YannickJadoul commented Dec 2, 2024

Several issues are currently breaking CI builds:

  • NumPy 2 has a different repr of scalars (e.g., np.float64(496.3505526721194) instead of 496.3505526721194)
    • Fixed in b5c5ad0 and 93d6817
    • Is 93d6817 a good fix? Is this what we want?
    • Python 3.8 doesn't have a NumPy >= 2 version. Python 3.8 is also EOL since a month or two. So we could just scrap it, though I prefer to keep that for minor updates (i.e., 0.3.0), and not for patch updates (0.2.2). So we probably better just ignore those two doctests on Python 3.8.
  • No abjad release supporting Python >= 3.12
    • I added 3.12 and 3.13 (04d0a88) cause we should probably be testing and supporting them, but so we cannot install abjad, currently.
    • Do we want to ping the abjad people? Do we want to find a way to get rid of the dependency? Do we just not want to test the abjad parts in CI on 3.12 and 3.13?
  • No lilypond binary Python package for Apple Silicon (ARM)

Misc changes:

  • nPVI tests weren't updated after fix. But easily fixed: 7fb0d43 (and checked independently that this is the right value)
  • Because of the shallow git checkout of the checkout action, tags and history weren't cloned. So setuptools_scm couldn't get the version number right, which confused me while tracking down other issues. Fixed in 7498a9f.

@YannickJadoul
Copy link
Collaborator Author

Python 3.8 doesn't have a NumPy >= 2 version. Python 3.8 is also EOL since a month or two. So we could just scrap it, though I prefer to keep that for minor updates (i.e., 0.3.0), and not for patch updates (0.2.2). So we probably better just ignore those two doctests on Python 3.8.

Fixed in b8f61a9, but:

  • Is there a better solution?
  • conftest.py in root or in thebeat/ (can't be in tests/, as it's not picked up for doctests there)
  • Check for Python version or check for NumPy version?

@YannickJadoul
Copy link
Collaborator Author

No lilypond binary Python package for Apple Silicon (ARM)

Backported cefa9d9 from #82. Is this what we want, sweeping things under the rug for now?

@YannickJadoul
Copy link
Collaborator Author

YannickJadoul commented Dec 12, 2024

Fixed in b8f61a9, but:

Drop current solution; just skip test for now, and reenable on main, while working toward 0.3.0

@YannickJadoul
Copy link
Collaborator Author

Same here, what to do? Drop this option? Add documentation? Just don't test? Or test back on a non-ARM image?

For now, keep as is, and don't test on mac ARM.
But check if there's a way of only installing the optional dependency on intel.

@YannickJadoul
Copy link
Collaborator Author

I added 3.12 and 3.13 (04d0a88) cause we should probably be testing and supporting them, but so we cannot install abjad, currently.

Pinged Abjad about this. Removing 3.12 and 3.13 CI configs for this PR, and adding them in a separate one!

@YannickJadoul YannickJadoul marked this pull request as ready for review December 12, 2024 12:02
…ating point value of the pvalue was slightly different on Apple Silicon)
@Jellevanderwerff Jellevanderwerff merged commit 99ca620 into stable Dec 12, 2024
35 checks passed
@YannickJadoul YannickJadoul deleted the fix-stable-ci branch December 12, 2024 12:52
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