Uh oh!
There was an error while loading. Please reload this page.
[3.9] bpo-43077: Update bundled pip to 21.0.1 and setuptools to 52.0.0 (GH-24386) - #24387
[3.9] bpo-43077: Update bundled pip to 21.0.1 and setuptools to 52.0.0 (GH-24386)#24387pfmoore wants to merge 1 commit into
Conversation
pythonGH-24386) Update bundled pip to 21.0.1 and setuptools to 52.0.0. (cherry picked from commit 4d11ecb) Co-authored-by: Paul Moore <p.f.moore@gmail.com>
pfmoore
commented
Jan 30, 2021
:-( Looks like ensurepip changed between 3.9 and 3.10, probably to handle the fact that we're no longer shipping |
| _PIP_VERSION = "20.2.3" | ||
| _PACKAGE_NAMES = ('setuptools', 'pip') |
There was a problem hiding this comment.
The PR didn't backport cleanly, and I had to manually fix the merge conflict. I didn't notice that this variable isn't used in the older version of the code. It's harmless, but can be fixed. It's not the cause of the test failures, sadly.
pfmoore
commented
Jan 30, 2021
Test failures seem to be coming from a deprecation warning for a "legacy version" where the version being parsed is "pip". I have no clue why that's happening, but modifying the test ( My tentative assumption is that something in the code has been incorrectly parsing "pip" as a version for some time, but it's been harmless until now. With legacy versions now being warned about, that causes the test to fail with a hard error here. The bad news is I have no idea what is causing the code to try to parse "pip" as if it were a version. It's almost impossible to instrument the code, because it's being run directly from the wheel, and if I modify the code I'll need to regenerate I don't have time right now to debug this much further. @pradyunsg if you have any insights, I'd be very grateful! |
pfmoore
commented
Jan 30, 2021
This PR is stale because it has been open for 30 days with no activity. |
sbidoul
commented
Apr 24, 2021
Superseded by #25578 |
Update bundled pip to 21.0.1 and setuptools to 52.0.0.
(cherry picked from commit 4d11ecb)
Co-authored-by: Paul Moore p.f.moore@gmail.com
https://bugs.python.org/issue43077