python312Packages.nose: fix, use pep517 builder #325935
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Closed
13 tasks
dotlambda
reviewed
Jul 9, 2024
jchv
force-pushed
the
fix/nose
branch
2 times, most recently
from
657e500
to
fa76fc6
Compare
jchv
changed the title
13 tasks
dotlambda
reviewed
Jul 9, 2024
|
Since we are not pulling the patches from upstream, what is their license? |
|
Oh, that's a good question. The answer is unfortunately unclear. It turns out the GitHub repository for aports is just a mirror, so I probably need to switch the fetchpatch2 URLs again. But that said, the license situation for aports is unclear. https://gitlab.alpinelinux.org/alpine/aports/-/issues/9074 So maybe we can't use Alpine Linux's patches :/ That said, the patches are fairly trivial. I was working on my own version of this before I pulled in the Alpine patches. I guess I'd better switch back to doing that? |
|
I went to their chat and asked, we'll hopefully receive an answer soon. |
|
We can assume the published patches to be derivative works of nose, and as such they would have to be published under (L)GPL. I think we can continue with this assumption. |
13 tasks
|
SGTM. Just in case, I did also post an alternate patch. #325968 If we're going ahead with this PR, I think I want to do the following:
|
ofborg
bot
requested review from
bjornfor,
jbaum98,
vrthra,
wolfgangwalther,
dali99,
evanjs,
jluttine and
natsukium
ofborg
bot
added
10.rebuild-darwin: 1001-2500
and removed
10.rebuild-darwin: 501-1000
labels
|
The 3.12 patch is the one authored by the pynose maintainer (or at least the metadata claims that authorship, which isn’t necessarily proof positive). So, uh, I guess at least we have confirmation that that one is intended to be LGPL now. The 3.11 one seems too trivial and mechanical to be eligible for copyright protection anyway. I do think that this seems like a better approach than pulling in further updates from the pynose upstream. |
|
This pull request has been mentioned on NixOS Discourse. There might be relevant details there: https://discourse.nixos.org/t/error-nose-1-3-7-not-supported-for-interpreter-python3-12/48703/17 |
|
FWIW, both of the patches are actually pretty trivial: I didn't look at it in great detail, but I believe the largest segment of code is the That said, if we have too much cold feet over it, I'm happy to pivot towards #325968 where I make a fully independent patch. It has the advantage that I started on it before finding the other patch, so it is legitimately different, although the truth is that they're going to be similar no matter what since it's the same idea. The only strong disadvantage is that I'm sure my patch is not complete yet. No issue for me either way. I've been running a nixpkgs-review run over night and all seems to be well with this patch by Michael Mintz. I haven't checked every single failure, but of them I have yet to see a failure that hasn't already existed on |
|
The pynose fork was eventually relicensed LGPL, so that part isn’t an issue. Learning that the |
|
Makes sense to me. I'm a bit busy but I think I should be able to find time today to rebase and clean up that PR so it is ready to merge. |
This comment was marked as outdated.
This comment was marked as outdated.
|
Result of 220 packages marked as broken and skipped:
326 packages failed to build:
2783 packages built:
|
|
Wow, pretty absurd timing. I wonder why my nixpkgs-review run had so many less failures. |
|
Closing in favor of #325968. |
|
Result of 220 packages marked as broken and skipped:
275 packages failed to build:
2834 packages built:
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description of changes
(Closed in favor of #325968 for now.)
This ports the Alpine Linux patches for Nose in Python 3.12 to Nixpkgs.
I don't know if I can get consensus on this as the "right" way to go, but with pynose having licensing issues, a ton of packages are broken right now. I don't personally think it makes sense for us to have a bunch of downstream ports of test suites to other test runners. It may be relatively easy, but these patches for Nose are also pretty easy, except they fix Nose for everything simultaneously and only need to be touched when Python updates rather than potentially when any Nose-dependent downstream package updates.
Some packages are still broken in 3.12 for other reasons, so I didn't bother to go around and change other derivations. I think it would make sense to stop the bleeding first, if we can.
Things done
nix.conf? (See Nix manual)sandbox = relaxedsandbox = truenix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage./result/bin/)Add a 👍 reaction to pull requests you find important.