-
Notifications
You must be signed in to change notification settings - Fork 1
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
force numpy>2.0 instead of deprecated oldest-supported-numpy during build #55
Conversation
@ntessore This is ready for a review and testing, but you need to approve the action as I have never commited to this repo before (assuming default security settings) |
Hi @sroet, thanks for this PR. Can we keep Python 3.7 and 3.8 working by using |
Hey @ntessore
I would strongly advice you to follow SPEC0/NEP29 and drop support for In the mean time I will see if there is a way to make your request work |
@ntessore, you would have to re-approve the tests again dd6c2df re-enables the same build environment as the current main branch for python <3.9 Be aware though, that it might lead to broken builds (already on the current main) as this pulls in |
Thanks @sroet! I have no strong feelings one way or the other about supporting Python 3.7 to 3.9, but this package is very minimalistic in its actual requirements. Are the wheels for 3.10+ in any way negatively affected by keeping the <=3.9 versions around? |
Understandable, just wanted to make you aware that there are standards that are followed to reduce maintenance burden if required. I agree that this isn't probably needed with the minimalistic requirements
No, as far as I'm aware (just a slightly harder to read build-requirements list) (the split is 3.9+ vs <=3.8 in this PR) If instead you meant about only supporting 3.10+ , then you would also change |
Ok, as long as it's just the one hoop to jump through, it seems fair to keep support as it is. If it becomes more than one hoop, or the wheels start breaking, we can change the support matrix. Thanks a lot for the PR, give me a thumbs up if you are happy for me to merge. I will release a new version soon after. |
👍 from me. Thanks for the swift response and for writing such a useful and lightweight wrapper for healpix! |
closes #54 (specifically if a release is done somewhat soon after merging this)
After toying a bit with the local installation, this change actually resolved the error.
apparently oldest-supported-numpy is deprecated
It shouldn't break installs as this is mentioned on the numpy docs nowadays:
Which wasn't the case before numpy
1.25
, and the original reason for the existence ofoldest-supported-numpy