-
-
Notifications
You must be signed in to change notification settings - Fork 17
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
Split tests into a separate package, plus rattler-build #107
base: main
Are you sure you want to change the base?
Conversation
…onda-forge-pinning 2024.12.24.08.00.50
…onda-forge-pinning 2024.12.24.08.00.50
Eliminate the recipe conditions for cuda_major, since only CUDA 12+ is supported now. Replace the jinja `{% if %}` conditions with inline selectors, since the latter are more compatible with different tools (e.g. conda-recipe-manager).
After moving the tests out of the main package, its size reduces to 1.5M. The tests package itself is 9M. Proposed in issue conda-forge#22.
…onda-forge-pinning 2024.12.24.08.00.50
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 ( I do have some suggestions for making it better though... For recipe/meta.yaml:
This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/12484953648. Examine the logs at this URL for more detail. |
Thanks I think reduced download sizes are definitely welcome! |
recipe/meta.yaml
Outdated
outputs: | ||
- name: torchvision | ||
- name: torchvision-tests | ||
test: |
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.
you might still need a requirements section no?
I feel like we have very few of these -tests
packages so it might be good to make this a good example!
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.
To be honest, I don't think so. This package has no build steps, so no build requirements are needed — at least conda-build handles it pretty much immediately:
Packaging torchvision-tests
WARNING: No files or script found for output torchvision-tests
number of files: 0
Fixing permissions
Packaged license file/s.
INFO :: Time taken to mark (prefix)
0 replacements in 0 files was 0.00 seconds
As for run dependencies — I don't think having any is correct either. After all, the package doesn't install any files.
So the all dependencies needed are test requirements, and these are listed.
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.
I think you are right. I just didn’t look at the failure
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.
Oh, I haven't looked at all the jobs but it's definitely failing because of the new output. To be honest, I'm not sure we should actually be publishing that extra package.
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.
I don’t know if we have the tooling to not publish it.
For pillow I used a giant if statement that had to manually be turned on. But I haven’t run the test in months due to that manual step
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.
Yeah, it's probably not worth making things more fragile over this. Should I file a request to allow this new output, or wait for other maintainers to look at it first?
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.
I think you can push through with the request.
My review should be fine
h-vetinari doesn’t mind multi output recipiez
I’m typically more apprehensive due to the increased complexion of debugging multi output recipies locally.
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.
Given the file size savings. I’m ok with this.
…onda-forge-pinning 2024.12.24.08.00.50
${{ 'skip test_url_is_accessible instead of hitting 20+ servers per run, since' if 0 }} | ||
${{ 'each server might be occasionally unresponsive and end up failing our CI' if 0 }} | ||
test_url_is_accessible | ||
${{ 'spurious failure because upstream skip (Image.__version__ >= "7") does not trigger for Pillow "10"' if 0 }} | ||
or (test_transforms and test_adjust_saturation) |
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.
This is basically the best I could come up with, for a way to intersperse strings and comments here. Of course, I'm open to better suggestions.
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.
Can’t review so easily on my phone. But my main goal was to make it easy to expand and add new tests one line at a time to make the diff obvious.
This seems like it follows that spirit!
Hi! This is the friendly automated conda-forge-linting service. I failed to even lint the recipe, probably because of a conda-smithy bug 😢. This likely indicates a problem in your This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/12495694887. Examine the logs at this URL for more detail. |
I've added a rattler-build conversion, and filed output request in conda-forge/admin-requests#1248. FWICS conda-smithy crashes on it, I'm going to look into that later today. |
…onda-forge-pinning 2024.12.24.08.00.50
I'm going to guess that the osx failure is a fluke. |
The conda-smithy problems are conda-forge/conda-smithy#2165 and conda-forge/conda-smithy#2196. The former I've been able to workaround by avoiding nested conditions. |
I came to check in, and saw
failed due to connection issues. I always feel bad since conda-forge hammers webservers which are meant to be cached locally for data. Should we skip this test? |
Checklist
0
(if the version changed)conda-smithy
(Use the phrase@conda-forge-admin, please rerender
in a comment in this PR for automated rerendering)