Add new M1 macOS GitHub runner to CI

[RMeli]

Add new M1 macOS runner to CI.

---

📚 Documentation preview 📚: https://mdanalysis--4442.org.readthedocs.build/en/4442/

[github-actions]

Linter Bot Results:

Hi @RMeli! Thanks for making this PR. We linted your code and found the following:

There are currently no issues detected! 🎉

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (6ba7682) 93.37% compared to head (9c23ba0) 93.37%.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #4442      +/-   ##
===========================================
- Coverage    93.37%   93.37%   -0.01%     
===========================================
  Files          171      185      +14     
  Lines        21737    22851    +1114     
  Branches      4012     4012              
===========================================
+ Hits         20298    21338    +1040     
- Misses         952     1025      +73     
- Partials       487      488       +1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[IAlibay]

would be worth enabling cron CI - I can't remember if non-wheel build will work here, you might need to exclude macos-14 & wheels=false

[RMeli]

#4442

[RMeli]

macos-14 jobs are still queuing... Maybe it's work testing just 3.12 in the cron job too?

[IAlibay]

@RMeli I think there's a runner limit, I might be the one using that up on my other PR. I can definitely run 4 concurrent runners there.

[IAlibay]

Actually I didn't see, these have been queing for 20h?! That looks like a problem with github actions itself. Wouldn't surprise me given the beta nature of these runners. Maybe we're jumping the gun too early?

[RMeli]

That looks like a problem with github actions itself. Wouldn't surprise me given the beta nature of these runners. Maybe we're jumping the gun too early?

Possibly a bit unstable, but I encountered not such problem in RMeli/spyrmsd#102. Maybe it's just that everyone rushed to test them as we did?

[RMeli]

All green, at last.

[IAlibay]

One blocking question.

P.S. the runner failure you were saying is down to the runs-on being incorrect - if it isn't recognise gh actions doesn' t fail but just goes on indeterminate hold.

[IAlibay]

why is this being disabled? as show in #4445 we should be able to build on 3.9 with macos-14

[RMeli]

Version 3.9 was not found in the local cache
Error: The version '3.9' with architecture 'arm64' was not found for macOS 14.2.1.
The list of all available versions can be found here: https://raw.githubusercontent.com/actions/python-versions/main/versions-manifest.json

I got this failure. Looking at the versions-manifest.json as suggested, it looks like all 3.9 is not provided with arch: arm64.

[IAlibay]

That's.. odd. I've not glanced upstream, but we should probably keep an eye on updates - it's very possible they couldn't backport stuff properly until they had their own CI runners for this.

[IAlibay]

Couple of extra things. Also please do update the changelog with your handle.

[IAlibay]

That's.. odd. I've not glanced upstream, but we should probably keep an eye on updates - it's very possible they couldn't backport stuff properly until they had their own CI runners for this.

[IAlibay]

cheers, lgtm!