Closed hmaarrfk closed 8 months 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.
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.
Hi! This is the friendly automated conda-forge-linting service.
I wanted to let you know that I linted all conda-recipes in your PR (recipe
) and found some lint.
Here's what I've got...
For recipe:
@conda-forge-admin please rerender
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.
This is mostly really useful for those that want to start testing with qt6
@conda-forge/opencv thoughts on maintainability are appreciated.
@conda-forge-admin please rerender
Sorry it seems like there is still some work to do on windows.
@conda-forge-admin please rerender
@conda-forge-admin please rerender
I'm not a maintainer, but overall LGTM, just some minor suggestions.
@conda-forge-admin please rerender
@conda-forge-admin please rerender
@conda-forge-admin please rerender
I'm not convinced this is the right approach due to increased maintenance burden.
@hmaarrfk What is the recommended approach for pypi
packages which depend on opencv-python-headless
? Others like Intel provide such a package https://anaconda.org/intel/opencv-python-headless .
What is the recommended approach for pypi packages which depend on opencv-python-headless
pip install --no-deps package
is my only recommendation.
I'm not sure. I feel like opencv-python-headless
is a workaround from the combined limitations of pypi and opencv. We have different limitations here which make me hesitant to merge something like this.
The main issue with pypi is that one package may depend on opencv-python
and the other on opencv-python-headless
.
This merge request (as it is) wouldn't make pypi find a locally installed version of opencv-python-headless
and thus would not resolve your issue.
I would accept updates to this feedstock that make it look like a opencv-python-headless
is installed along side with opencv-python
.
Perhaps you can be inspired by:
https://github.com/conda-forge/opencv-feedstock/blob/main/recipe/install_pip_metadata.patch
I would ask that you add alot of skips
when developing on the CIs for now. The build matrix for opencv is already quite huge.
Hi @hmaarrfk, thank you for this piece of work!
I might be able to help with packaging the equivalent of opencv-headless-python
.
Is there a lot of work left? Would you be able to estimate the time it would take to finish it? Can I contribute to your efforts?
on the coding side, the answer is that it is "done".
the question I have is:
if you have ideas on how to deliver it to the wider audience I am glad to hear it.
personally, I see this request as quite niche, and those that require it can likely rebuild opencv using azure's free agents and upload the package to their channel. the package would be called opencv
and simply not install the qt aspects for users that subscribe to your channel.
Checklist
0
(if the version changed)conda-smithy
(Use the phrase code>@<space/conda-forge-admin, please rerender in a comment in this PR for automated rerendering)