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
python.pkgs.dm-sonnet: 1.30 -> 1.33 #63809
Conversation
@FRidh continuing the discussion from #63806 (comment): Do you have any idea why pip fails to find any version of |
The package probably did not create/provide a proper dist-utils folder. |
You're right, there is no distutils folder. I noticed that upstream installation instructions of Any idea why? How is that folder usually created? |
Note it is actually the In the past I've encountered similar issues where upstream was using some other kind of build-system, thus not creating that folder. At the time they also provided a In your case I suppose |
Yes, I can use git clone https://github.com/tensorflow/probability.git
cd probability
bazel build --copt=-O3 --copt=-march=native :pip_pkg
PKGDIR=$(mktemp -d)
./bazel-bin/pip_pkg $PKGDIR
pip install --user --upgrade $PKGDIR/*.whl But doing that,
If I remove
So the wheel is created, but the .dist-info dir apparently not. Also the part where it tries to replace Here's a link to their setup.py: https://github.com/tensorflow/probability/blob/master/setup.py |
That's odd considering the README explicitly mentions that tensorflow is not included. Anyway, you could consider passing Note I'll be away for roughly a week or two. |
Ah, I fixed the "no dist-info" problem. I unpacked the built wheel and noticed that the package is called Unfortunately to fix it I had to patch the setup.py, because as far as I can see there is no way to pass flags to The attempt to install |
@GrahamcOfBorg build python3.pkgs.dm-sonnet |
This is good to go from my side. @kalbasit do you want to take a look at the changes to |
To pass a flag to |
Not sure how I could've missed Thank you for your help! I hope you'll have a nice vacation or wherever you're off to. |
@kalbasit do you have any package you want to test the buildBazelPackage normalization changes on? I'm not entirely sure what the normalization is intended to do and wasn't able to find any examples. |
Previously the installPhase of the fixed ouput derivation would fail for a package that has no markers, since `sed` would complain about having no input files. If we use `find` instead of bash globs, that problem goes away.
Fixes the tensorflow-probability built by first building the wheel with bazel. This actually creates the dist-info folder, allowing the package to be picked up as a pip dependency.
I've mostly used it for packaging bazel-watcher, let's try it out. @GrahamcOfBorg build bazel-watcher |
ofBorg gave up, but I confirmed locally that it builds and the fixed output hash remains the same. So good to go from you @kalbasit? |
|
@timokau yes that's fine. I'll give it another shot, just in case. @GrahamcOfBorg build bazel-watcher |
@timokau LGTM. Please merge at your earliest convenience. |
Thank you for taking a look! |
Motivation for this change
Things done
sandbox
innix.conf
on non-NixOS)nix-shell -p nix-review --run "nix-review wip"
./result/bin/
)nix path-info -S
before and after)