Closed AbdealiLoKo closed 1 year ago
Let me mull it over, the 2014 is expected to be EOL in January of next year. My feeling at the moment is that it'd probably be easier for you to fork pysqlite3
and register pysqlite3-binary-manylinux2014
or something on pypi and push these alternate wheels there.
Thanks @coleifer I have my own internal wheelhouse that I am already building. Not sure if I'll publish to pypi, but will be maintaining it internally for my machines running on amazonlinux
On a side note, wanted to mention that manylinux2010 will be end of life after July 2024: https://github.com/mayeut/pep600_compliance#acceptable-distros-to-build-wheels
The pypa/manylinux readme, which I consider to be an "official" source, declares that manylinux2010 was end-of-life last year:
For now I'm going to hold off on merging this change.
@coleifer I think there is some confusion ...
I agree that 2010 is EOL. This PR was for 2014 which is not EOL
Sorry yes, I was mixed up in my last comment. You are right about 2014 having a bit over a year left before EOL.
Nonetheless, and I'm sorry because I see the merit of your request, I'm going to hold off for the moment on any changes.
Hey there, following up on https://github.com/coleifer/pysqlite3/issues/40#issuecomment-1523882073
I went through your code, and I realized the structure was super intuitive :+1: And before I knew it I was able to get the project working on manyinux_2_17 quite easily :)
So, decided to create a PR to use the manylinux2014 (aka manylinux_2_17) docker from pypa.
I tested this locally, and:
If this seems good, would be great to get this merged !