Example of pkgconf usage

[conda-forge-admin]

Hi! This is the friendly automated conda-forge-webservice.

I've started rerendering the recipe as instructed in #316.

If I find any needed changes to the recipe, I'll push them to this PR shortly. Thank you for waiting!

Here's a checklist to do before merging.

• Bump the build number if needed

[conda-forge-admin]

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/meta.yaml) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe/meta.yaml:

• ℹ️ The recipe is not parsable by parser conda-souschef (grayskull). This parser is not currently used by conda-forge, but may be in the future. We are collecting information to see which recipes are compatible with grayskull.
• ℹ️ The recipe is not parsable by parser conda-recipe-manager. The recipe can only be automatically migrated to the new v1 format if it is parseable by conda-recipe-manager.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/13534727439. Examine the logs at this URL for more detail.}

[hmaarrfk]

@conda-forge-admin please rerender

[hmaarrfk]

@traversaro sorry for the ping. But it seems that ffplay is not compiling anymore.

Do you mind taking a look?

is there a way to check if it is enabled at build time instead of test time

[traversaro]

@traversaro sorry for the ping. But it seems that ffplay is not compiling anymore.

Do you mind taking a look?

is there a way to check if it is enabled at build time instead of test time

No problem, I will check later this (european) evening.

[traversaro]

It could be a regression due to the update of sdl2 package to sdl2-compat.

[hmaarrfk]

do you have a link to the feedstock?

[traversaro]

do you have a link to the feedstock?

conda-forge/sdl-feedstock#21

[hmaarrfk]

it does seem related since pinning to 2.30.10 seems to working, i'll let you debug this if thats ok....

[traversaro]

it does seem related since pinning to 2.30.10 seems to working, i'll let you debug this if thats ok....

Sure.

[traversaro]

If you want to proceed without being blocked as the later version are ABI compatible pinning the host sdl2 to an earlier version is perfectly fine.

[hmaarrfk]

Thanks. That’s what I did just now. Rebased this branch on the newly merged main

[traversaro]

I did not double checked if this is the case, but I guess the problem could be related to conda-forge/sdl-feedstock#22 .

[hmaarrfk]

shall we try pkgconf here?

[traversaro]

shall we try pkgconf here?

To be honest I am not sure if we want to go in the rabbit hole of understanding how to tell ffmpeg to use pkgconf instead of pkg-config (see discussion in conda-forge/conda-forge.github.io#1880 (comment)). I implemented a solution compatible with old pkg-config in conda-forge/sdl-feedstock#23 , hopefully it will fix the problem.

[traversaro]

understanding how to tell ffmpeg to use pkgconf instead of pkg-config

shouldn't it be as simple as just specifying the path to the binary? If we want, we don't even need to rely on anything from ffmpeg for that, we could just use pkgconf and create a symlink to it from $PREFIX/bin/pkg-config.

Yes, and personally I would avoid doing that for all the feedstock that use pkg-config with sdl2. conda-forge/sdl-feedstock#23 should fix the issue, if that does not work I will look into it.

[hmaarrfk]

Yes, and personally I would avoid doing that for al

if nobody experiments with pkgconf we will be doomed with pig-config for ever.

I think this is a really good opportunity to do so….

[traversaro]

Yes, and personally I would avoid doing that for al

if nobody experiments with pkgconf we will be doomed with pig-config for ever.

I think this is a really good opportunity to do so….

Probably this is a bit OT, but I think it make sense to experiment with pkgconf, I would just experiment with it exactly in the way all the distros I am aware of (see list in the following) experimented/worked on it, that is by providing an alternative "pkg-config" package that provides a pkg-config executable that actually uses pkgconf under the hood. Anyhow, this is just a personal preference, anyone is welcome to work on any alternative path or experiment.

Example of switch from pkgconf to pkg-config:

• Debian: https://groups.google.com/g/linux.debian.devel/c/kmS6iIcClPY
• Fedora: https://fedoraproject.org/wiki/Changes/pkgconf_as_system_pkg-config_implementation
• vcpkg: [vcpkg_find_acquire_program(PKGCONFIG)] Switch to pkgconf, add arm64 microsoft/vcpkg#31924

[hmaarrfk]

I think conda-forge is very unique in the heavy use of cross compilation and simultaneous support Linux + OSX + Windows.

if anything were to accelerate the adoption of pkgconf, it would be "the ffmpeg used the compatibility layers we are proposing to rollout conda-forge wide". Otherwise the "conda-forge existance proof is missing".

I think the only missing piece would be a package that uses pkg-config natively for windows????

(I just rebased and rerendered force push FYI)

[traversaro]

if anything were to accelerate the adoption of pkgconf, it would be "the ffmpeg used the compatibility layers we are proposing to rollout conda-forge wide". Otherwise the "conda-forge existance proof is missing".

Sure, but to be honest I am not super-familiar with messing with the host files installed by other packages, that is why I would prefer to avoid that. It seems to me (but again, this is just my opinion, I am not stopping anyone to work on this) that starting with having a dual build of pkg-config package, the proper pkg-config one with higher priority/build number, and one with a lower priority with pkgconf, and the use the pkgconf one here (or in another repo that uses pkg-config) would be more sound as a way to test the switch to pkgconf.

[hmaarrfk]

@traversaro I think something like this is 'reasonable' while we know for a fact that it is worthwhile to create the conflict between packages.

Already we have revealed a usecase where cairo.h cannot be found??? Not sure if that is the actual error or an other problem.

[traversaro]

Yes, I think this make sense. One of my doubts was how to ensure that the symlink did not resulted in a file that is part of the new package, but the trick of PKGCONF_LINKED + rm fixes that.

[h-vetinari]

Yes, and personally I would avoid doing that for all the feedstock that use pkg-config with sdl2.

Are you saying that sdl2 depends on bugs that are specific to pkg-config. Because aside from bugs, pkgconf should do exactly the same, so I'm confused what conditions would lead to a genuine need to avoid pkgconf.

[hmaarrfk]

@conda-forge-admin please restart cis

[hmaarrfk]

Windows is still failing but we are closer on linux

[traversaro]

Yes, and personally I would avoid doing that for all the feedstock that use pkg-config with sdl2.

Are you saying that sdl2 depends on bugs that are specific to pkg-config. Because aside from bugs, pkgconf should do exactly the same, so I'm confused what conditions would lead to a genuine need to avoid pkgconf.

I was referring to the fact that I would avoid having to manually specify pkgconf in any package that depend on sdl2, not that pkgconf should be avoid in general.

[traversaro]

Windows is still failing but we are closer on linux

It seems that the latest failing build was not using the build _1 done by conda-forge/sdl-feedstock#23, that is required for both pkgconf and pkg-config on Windows as no .pc file was installed at all before it.

[hmaarrfk]

sweet lets leave this as an pkgconf example, and do the cleanup in #319