-
-
Notifications
You must be signed in to change notification settings - Fork 46
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
Backport run_exports #320
Comments
Not sure what you're asking here. Numpy in conda-forge has had run-exports since 1.19, but the compatibility (i.e. the version constraints in the numpy run-export) changed fundamentally with numpy 2.0, and cannot be backported. |
Could you please share more about the issue you are encountering? |
Oh, okay, I just assumed that abi compatibility was for a while. In other words I was assuming if project was compiled with any python version 1.19 - 1.26 it will be compatible with |
No NumPy didn't introduce the new ABI compatibility until NumPy 1.25.0 We discussed whether we wanted to change things then ( conda-forge/conda-forge-pinning-feedstock#4816 ). However we decided it wasn't worth the effort as we could do this with NumPy 2 (when we already would need to change things) So the approach we are taking is rebuild everything with NumPy 2 (as we need to anyways) and use the new ABI there. The packages produced work on NumPy 1 & 2, which is a bigger benefit for users then slightly more compatibility with NumPy 1 versions (and no NumPy 2 support) |
Yes, that totally makes sense! Thank you for explaining! |
Comment:
Is it possible to backport changes in #313 ? It would be nice to have it numpy 1.19 - 1.26
The text was updated successfully, but these errors were encountered: