search

conda-forge / scipy-feedstock

A conda-smithy repository for scipy.
BSD 3-Clause "New" or "Revised" License
6 stars 36 forks source link

SciPy 1.11.1 #243

Closed h-vetinari closed 1 year ago

h-vetinari commented 1 year ago

A quick release to fix a licensing issue (scipy actually yanked 1.11.0; not sure if we want to mark the builds as broken)

conda-forge-webservices[bot] commented 1 year ago

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

h-vetinari commented 1 year ago

From a quick calculation based on this, we've already had >50k downloads.

rgommers commented 1 year ago

When you mark a build as broken, will it still be installable when the exact version is contained in, for example, a lock file produced by conda-lock?

h-vetinari commented 1 year ago

No, the build is then truly unavailable and will break reproducibility of any environment that got created in the meantime. This works differently than the yanking mechanism on PyPI.

There is a way to reproduce the environment, but I guess it's not so widely known, which is to add a special channel -c conda-forge/label/broken, then the environment can be reproduced.

But it's a big hammer, and I honestly doubt it's worth it

rgommers commented 1 year ago

But it's a big hammer, and I honestly doubt it's worth it

Agreed I think (unless there's a conda-forge policy that says the opposite) - it was yanked because of one file that shouldn't been removed, and that file wasn't even used in the build.

isuruf commented 1 year ago

When you mark a build as broken, will it still be installable when the exact version is contained in, for example, a lock file produced by conda-lock?

Yes, lock files will continue to work.

No, the build is then truly unavailable and will break reproducibility of any environment that got created in the meantime.

Not true.

This works differently than the yanking mechanism on PyPI.

Kind of. Lock files will keep working, however you can't install it with == unlike PyPI.

h-vetinari commented 1 year ago

Thanks for the clarification. I was not aware that lock files will continue to work.