try qt6-main - Githubissues

looooo commented 9 months ago

Checklist

[ ] Used a personal fork of the feedstock to propose changes
[ ] Bumped the build number (if the version is unchanged)
[ ] Reset the build number to 0 (if the version changed)
[ ] Re-rendered with the latest conda-smithy (Use the phrase code>@<space/conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
[ ] Ensured the license file is being packaged.

conda-forge-webservices[bot] commented 9 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.

looooo commented 9 months ago

for osx-arm64 add this to the build-script:

if [[ "${target_platform}" =~ osx-arm64 ]]; then
    rm -f "${PREFIX}/lib/qt6/moc"
    ln -s "${BUILD_PREFIX}/lib/qt6/moc" "${PREFIX}/lib/qt6/moc"

    # Additional debugging information
    echo "Adjusted Qt tools for osx-arm64 with build variant qt6"
    echo "Removed: ${PREFIX}/lib/qt6/moc"
    echo "Linked to: ${BUILD_PREFIX}/lib/qt6/moc"
else
    echo "Skipping Qt tools adjustment. Target platform: ${target_platform}, Build variant: $build_variant"
fi

osx-64:

ld: framework not found Metal

looooo commented 9 months ago

@conda-forge/core build: -skipis here for me not working. any ideas? I would like to keep the build matrix small and only build for py>311 additional the osmesa and egg variants are not needed as these builds are not qt dependent (TODO).

also I would like to know if we can upload with the label qt6 (currently it is set to vtk_qt6) but I guess having a qt6 label is more convenient as this can be used for more packages...

hmaarrfk commented 9 months ago

@conda-forge-admin please rerender

conda-forge-webservices[bot] commented 9 months ago

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

I was trying to look for recipes to lint for you, but it appears we have a merge conflict. Please try to merge or rebase with the base branch to resolve this conflict.

Please ping the 'conda-forge/core' team (using the @ notation in a comment) if you believe this is a bug.

hmaarrfk commented 9 months ago

can you rebase and remove the rerendering comits

conda-forge-webservices[bot] commented 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.

looooo commented 8 months ago

@hmaarrfk what do you think about adding a qt6 label to the packages and use a new branch?

hmaarrfk commented 8 months ago

why? just rip off the bandaid tbh, it is very confusing to constantly tell people that they need to use a specific label or branch.

If you want to tell people to use a specific label, then you might as well just build our own channel (essentially equivalent) and keep it as updated as you want. It works on a small scale.

I think a better solution would be to update to 9.3 and then users can pin to to 9.2 for "qt5" and "9.3" for qt6.

Qt6, with the exception of qt-webengine, has been doing quite well.

looooo commented 8 months ago

why? just rip off the bandaid tbh, it is very confusing to constantly tell people that they need to use a specific label or branch.

I just wanted to not introduce troubles for packages depending on qt5 and vtk. But I am happy to have it directly in conda-forge.

I think a better solution would be to update to 9.3 and then users can pin to to 9.2 for "qt5" and "9.3" for qt6.

sounds good to me.

hmaarrfk commented 8 months ago

I don't really know a way to avoid the troubles.

My attempt is to make an easy version cutoff that people that need old can still use.

It is really hard to use build strings for this since build strings are somewhat adhoc

conda-forge-webservices[bot] commented 8 months ago

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:

Selectors are suggested to take a <two spaces>#<one space>[<expression>] form. See lines [34, 42, 209, 235]

conda-forge-webservices[bot] commented 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.

looooo commented 8 months ago

The reported errors are:
- Encountered problems while solving:
-   - package vtk-base-9.3.0-osmesa_py312h1234567_100 requires libxml2 >=2.11.6,<2.12.0a0, but none of the providers can be installed

hmaarrfk commented 8 months ago

I've had alot of trouble with the libxml2122 migration https://github.com/conda-forge/qt-main-feedstock/pull/216/files

I've been going through the graph and doing it myself https://conda-forge.org/status/#libxml2212

it seems that it is now vtk's turn.

looooo commented 7 months ago

@conda-forge-admin, please rerender

looooo commented 7 months ago

I've had alot of trouble with the libxml2122 migration https://github.com/conda-forge/qt-main-feedstock/pull/216/files

I've been going through the graph and doing it myself https://conda-forge.org/status/#libxml2212

it seems that it is now vtk's turn.

seems like this pr is using libxml2 2.12.4, the pinning is:

libxml2:
- '2'

the failing build has error messages like this:

/home/conda/feedstock_root/build_artifacts/vtk_1706771685165/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placeho/include/GL/osmesa.h:323:1: error: expected '=', ',', ';', 'asm' or '__attribute__' before 'OSMesaPostprocess'

hmaarrfk commented 7 months ago

we loosed libxml2 due to the fact that we believe it now better maintains abi compatibility accross v2.

It is unclear from the short error message that you have that the error is related to the libxml2.

looooo commented 6 months ago

we loosed libxml2 due to the fact that we believe it now better maintains abi compatibility accross v2.

thanks, good to know.

Any ideas what to do with the failing build? If not, is it possible to merge also with the failing build?

looooo commented 6 months ago

@conda-forge-admin, please rerender

hmaarrfk commented 6 months ago

it seems that the failing build is a mesalib build.

was it working before? mesalib 24 just got released, so maybe there is some incompatibility.

adrianinsaval commented 6 months ago

@looooo could you check your personal repo's setting? it's uploading stuff to the freecad repo in anaconda which then breaks the freecad weekly builds https://cirrus-ci.com/task/4937101846773760?logs=osx_build#L1012

conda-forge-webservices[bot] commented 3 months ago

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:

The feedstock has no .ci_support files and thus will not build any packages.

For recipe:

You're setting a constraint on the __osx virtual package directly; this should now be done by adding a build dependence on {{ stdlib("c") }}, and overriding c_stdlib_version in recipe/conda_build_config.yaml for the respective platform as necessary. For further details, please see https://github.com/conda-forge/conda-forge.github.io/issues/2102.
In your conda_build_config.yaml, please change the name of MACOSX_DEPLOYMENT_TARGET, to c_stdlib_version!

looooo commented 3 months ago

@conda-forge-admin, please rerender

github-actions[bot] commented 3 months ago

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

I tried to rerender for you but ran into some issues. Please check the output logs of the latest webservices GitHub actions workflow run for errors. You can also ping conda-forge/core for further assistance or you can try rerendeing locally.

This message was generated by GitHub actions workflow run https://github.com/conda-forge/vtk-feedstock/actions/runs/9331566344.

conda-forge-webservices[bot] commented 3 months ago

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:

The feedstock has no .ci_support files and thus will not build any packages.

For recipe:

In your conda_build_config.yaml, please change the name of MACOSX_DEPLOYMENT_TARGET, to c_stdlib_version!

looooo commented 3 months ago

@conda-forge-admin, please rerender

github-actions[bot] commented 3 months ago

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

I tried to rerender for you but ran into some issues. Please check the output logs of the latest webservices GitHub actions workflow run for errors. You can also ping conda-forge/core for further assistance or you can try rerendeing locally.

This message was generated by GitHub actions workflow run https://github.com/conda-forge/vtk-feedstock/actions/runs/9331856024.

looooo commented 3 months ago

@conda-forge/core is this rerendering issue known?

looooo commented 3 months ago

@conda-forge-admin, please rerender

github-actions[bot] commented 3 months ago

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

I tried to rerender for you but ran into some issues. Please check the output logs of the latest webservices GitHub actions workflow run for errors. You can also ping conda-forge/core for further assistance or you can try rerendeing locally.

This message was generated by GitHub actions workflow run https://github.com/conda-forge/vtk-feedstock/actions/runs/9338304061.

conda-forge-webservices[bot] commented 3 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.

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

For recipe:

In your conda_build_config.yaml, please change the name of MACOSX_DEPLOYMENT_TARGET, to c_stdlib_version!

conda-forge-webservices[bot] commented 3 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.

looooo commented 3 months ago

ppc64le fails with:

Encountered problems while solving:
  - nothing provides requested proj 9.4.0.*

hmaarrfk commented 3 months ago

maybe you have to skip it: https://github.com/conda-forge/proj.4-feedstock/pull/143

hmaarrfk commented 3 months ago

out of curiosity, why do you not want to build ffmpeg 6?

looooo commented 3 months ago

out of curiosity, why do you not want to build ffmpeg 6?

keep the build-matrix smaller. My goal is to provide freecad builds for qt6. I think it's best to do this first via the anaconda.org freecad organization. I have the fear that qt6 is making things a little bit messy. Or is there a proposal how to deal with qt6 without introducing too much issues for conda-forge?

hmaarrfk commented 3 months ago

you can skip it if you want. but ffmpeg7 is quite new, and many projects don't always support it right off the bat.

you can add a skip: true # [ffmpeg == '6'] or something like that and rerender.

or just build the two and forget about it ;)

looooo commented 3 months ago

@hmaarrfk I just looked into setting up azure for my personal repo, but I am a bit lost (everything seems to have changed there). What do you think about setting up a qt6 branch for the vtk-feedstock and publish the builds to a qt6-label?

hmaarrfk commented 3 months ago

I would not setup a branch unless it has a different version number for VTK.

It increases the chance of a clash.

I've been using qt6 for a while now, i think it is OK.

i would keep it simple for yourself.

Azure kinda broke my personal repo workflow, need time to get it back up and running.

looooo commented 3 months ago

@hmaarrfk so there is nothing special needed for qt6 (for example renaming the builds as we did here: https://github.com/conda-forge/soqt-feedstock/blob/main/recipe/meta.yaml#L1L8)?

hmaarrfk commented 3 months ago

you added a constraint since the two packages clash:

https://github.com/conda-forge/soqt-feedstock/blob/main/recipe/meta.yaml#L40

OpenCV currently builds 3 variants:

qt5
qt6
noqt

I'm trying to get rid of qt5 given the new version bump (easier to communicate) https://github.com/conda-forge/opencv-feedstock/pull/417

so given that you have a version bump here, might as well only build for qt6 and move on.

looooo commented 3 months ago

the osmesa build fails with:

_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placeho/include/GL/osmesa.h:323:1: error: expected '=', ',', ';', 'asm' or '__attribute__' before 'OSMesaPostprocess'

/**
 * Enable/disable Gallium post-process filters.
 * This should be called after a context is created, but before it is
 * made current for the first time.  After a context has been made
 * current, this function has no effect.
 * If the enable_value param is zero, the filter is disabled.  Otherwise
 * the filter is enabled, and the value may control the filter's quality.
 * New in Mesa 10.0
 */
GLAPI void APIENTRY
OSMesaPostprocess(OSMesaContext osmesa, const char *filter,
                  unsigned enable_value);

is there something obvious?

looooo commented 3 months ago

@hmaarrfk is it ok to skip the osmesa builds for now?

hmaarrfk commented 3 months ago

That's between you and the other maintainers.

Give them a week or two to respond or something?

looooo commented 3 months ago

continued in #328

conda-forge / vtk-feedstock

try qt6-main #320