-
Notifications
You must be signed in to change notification settings - Fork 319
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
Drop older Python versions #827
Conversation
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## master #827 +/- ##
========================================
- Coverage 38% 12% -27%
========================================
Files 88 88
Lines 13084 10628 -2456
Branches 1830 1331 -499
========================================
- Hits 5076 1340 -3736
- Misses 7583 9160 +1577
+ Partials 425 128 -297
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
I'd suggest to squash the commits |
at least those related to version removing. |
FYI, regarding commit squashing: we've been using squash merges for some time now in this project. Depending on how you want to organize your work the PR-local commit noise may not necessarily matter much. |
You are right, no problem, let's use squash merge. |
Yes I agree about the squash merge actually, it's a bit easier for the contributors. |
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.
LGTM
py{27,35,36,py2,py3}-epolls | ||
py{37,38,39,310}-{selects,poll,epolls} | ||
pypy3-epolls | ||
py{38,39,310}-{selects,poll,epolls} |
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.
Not sure why the deps of below testenv
manage Python 3.11 and why the above envlist
, apparently, not. As the semantic of these changes is to drop EOLed Python releases, this patch LGTM.
@temoto, @jstasiak: Even if I'm now an eventlet orga team member, I'm not able to merge patches from the eventlet repo. I think you need to tweak some specific rights of this repo. However, in a first time I'll only focus on easy patches like this one, to become more familiar with the eventlet repo and avoid mistake. |
Looks better, thanks. @itamarst: However, we should notice that the merge is still disabled due to the required tests that are failing, so what do you think of my development branch proposal made there? Using this development branch we could temporarly disable failing tests, then redirect this pull request and your other one (#823) against the development branch rather than against the master branch, then, merge them, and finally re-enabling all the tests. That would avoid disabling the tests, on all the master pull requests. Even if they tests are not functional, they protect us from merging patches accidentally. Patches merged on |
@jstasiak: I think the same admin action should be done for all the persons recently added. Big thanks for your time and for your help! |
Everyone else was already in the developers team so I don't think anything to do there. Let me know if it turns out otherwise though. FYI for your convenience I temporarily disabled the requirement for PRs to have green builds to be merged. Let me know if you want it undone (and when). |
Great, thanks |
@itamarst: which pull request are you desire to be merged first?:
|
My plan is:
|
I took a look at recent download stats from https://pypistats.org/packages/eventlet (and used its API to get raw stats so I could compute monthly aggregates) -- I worry that dropping 3.7 may have been a little hasty. While 3.7 downloads have been trending down for months, we've only just this month begun getting more downloads from 3.7 than 3.11; 3.7 accounts for roughly one in eight downloads.
Honestly, I'm not sure why there would be so many 3.7 downloads; my recollection was that Ubuntu LTS releases used 2.7, 3.8, 3.10, while RHEL/CentOS hit 2.7, 3.6, 3.9... DistroWatch gives me a hint that it may be Debian Buster? If so, at least their long term support would end come the middle of next year... The question seems particularly important, though, given the premise of eventlet as critical to whole ecosystems of projects and our need to address security issues. |
@thomasgoirand o/ Any opinion about the previous comment? |
I opened a new issue for the 3.7 discussion: #846 |
Fixes #819
Fixes #822
Tests won't pass for now, but should at least run as before, just fewer builders.