add stdlib metapackages to pinning #5499
Conversation
|
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 ( |
|
OK, it should be impossible to break something here, so putting this in to continue working on the various bits necessary for this. |
| c_stdlib: | ||
| - sysroot # [linux] | ||
| - macosx_deployment_target # [osx] | ||
| - vs # [win] | ||
| c_stdlib_version: # [unix] | ||
| - 2.12 # [linux64] | ||
| - 2.17 # [aarch64 or ppc64le] | ||
| - 10.9 # [osx and x86_64] | ||
| - 11.0 # [osx and arm64] |
There was a problem hiding this comment.
The macOS keys appear not to be in order
There was a problem hiding this comment.
Or should there be a Windows version?
There was a problem hiding this comment.
No, windows doesn't set a version, similar to how the windows compilers don't have c_compiler_version
| - # [unix] | ||
| - c_stdlib # [unix] | ||
| - c_stdlib_version # [unix] |
There was a problem hiding this comment.
While this appears to be its own group, it looks like the bot might be lumping them together. Not sure why that is
Taking this from a recent bot migration run:
ValueError: Variant configuration errors in /home/runner/.cache/conda-smithy/conda_build_config.yaml:
zip fields in zip_key group frozenset({'cuda_compiler_version', 'fortran_compiler_version', 'cxx_compiler_version', 'cuda_compiler', 'c_compiler_version', 'docker_image', 'cdt_name', 'c_stdlib_version', 'c_stdlib'}) are not all the same lengthThere was a problem hiding this comment.
Actually this just happens when re-rendering a feedstock: conda-forge/ucx-split-feedstock#158 (comment)
There was a problem hiding this comment.
Am thinking we may need to fold this into the zip_keys group(s) above instead of having it separate
There was a problem hiding this comment.
Oh am now seeing this did happen: #5592
Somehow got an old diff
Based on infrastructure from conda/conda-build#4981, using packages from
This is in the larger context of moving our baseline on MacOS to 10.13 as soon as feasible and to glibc 2.17 on linux later this year.
In particular, we should add this to be able to properly test the new setup in https://github.com/regro/conda-forge-feedstock-check-solvable, which got brought up in the context of regro/cf-scripts#2135.