-
Notifications
You must be signed in to change notification settings - Fork 134
Add Python 3.14 to workflows #1521
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
Conversation
|
nice! if the continue-on-error is on the step instead of on the job you get a green check. good work figuring that out! |
fgregg
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A real improvement over what we have now!
Not a perfect solution since it is showing the individual 3.14 jobs as green, when they should be red, but much better than my attempt which showed the whole job as red, when it should be green. I think this the best we can get until something like "allow_failure" gets implemented by github.
|
@bmos if we change |
|
@bmos, reading the code more for setup-actions, i don't think a hashed-commit pin is will also allows us to just use "3.14-dev" and have that get is the most recent prelease version. happy to have that previous behavior reverted. |
Comment out python 3.14 classifier
add `--ignore-requires-python` flag to `uv install -e. [all]` in pytest job
revert added uv flag
https://github.com/move-coop/parsons/actions/runs/16249114550/job/45876284578#step:5:12 |
it works locally (but without uv) i’ll see if i can get it work with uv |
|
okay, this indeed a behavior difference between uv and pip. however uv does not respect upper bounds for python requires anyway, so i think no action is really needed here. maybe a comment next to the matrix about that fact could be helpful, |
fgregg
left a comment
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
thanks for this @bmos. i think this is is a good state. if you wanted to bring back the hash-pinning, that would be fine, but that's up to you!
Per the discussion in #1515, I have added the latest build of python 3.14 to the testing matrix.
It is set to not fail when the tests fail, so it shouldn't confuse contributors but will allow us to keep an eye on what failures block support.
From running this, we can see some existing compatibility issues that will need to be addressed before supporting 3.14:
lxml(v6.0.0) was included becauseparsons(v5.2.0) depends onsimple-salesforce(v1.12.6) which depends onzeep(v4.3.1) which depends onlxml]psycopg2-binary(v2.9.10) was included becauseparsons(v5.2.0) depends onpsycopg2-binary]There may be other issues after these are resolved. Both packages are the latest versions, which indicates we are most likely dependent on upstream fixes.