Conversation
Fixes archlinux CI
|
Documentation preview for this PR (built with commit aca1905; changes) is ready! 🎉 |
|
are we supporting 3.13 well enough? |
From a quick look I don't see any CI failure that could be blamed on this (didn't check all of them since there are tons of failing jobs due to multiple random issues) |
I checked that the arch CI job did run with python 3.13.1 and went well module some test failures, most of them not relevant to this PR. |
sagemathgh-39381: Allow system python 3.13 Fixes archlinux CI URL: sagemath#39381 Reported by: Antonio Rojas Reviewer(s): Dima Pasechnik
|
This has broken building on OS X + homebrew, as reported by several people on sage-release. Please back it out and implement it (and test it!) properly. |
detailed error reports welcome. I don't have your hardware and software |
|
I'm using OS X 15.3.1 on an Apple M1 machine. Homebrew packages: I'm attaching the top-level I built with a fresh copy of 10.6.beta7 + #39570. I used then the build succeeds. |
I think this is fixed by #39571 Could you please try #39571 with python3.13 (and 3.12?) Sorry, it has not crossed my mind that you didn't take #39571 - as I was telling other people in this thread on sage-release that the latter is needed to fix the latest beta. |
Thank you, #39571 fixes the build problems with In detail: and and |
|
@jhpalmieri - this looks like an issue with ipython, probably fixable by updating |
|
Updating |
|
rings/integer.pyx have this test dependent on pari, and I don't get this error with #38749 (upgrade pari to 2.17) |
I don't get this with Python 3.12, only 3.13. Any ideas why? (I built separately, using the same PRs and changes (but not yet #38749) — all of the ones necessary to build with 3.13 plus the PR making giac optional. I then used Python 3.12 with one build, 3.13 with the other.) |
Fixes archlinux CI