The generated CI files have Python lint steps. On GitHub, the lint steps will fail the pipeline if there are any errors. On GitLab, they only warn and the pipeline continues. The behavior is inconsistent. Our pure Python packages fail the pipeline on lint errors, so I think the GitHub rule would be preferable here. If some repos really need to skip linting, I'd rather have a way to opt out for the misbehaved ones than letting all the good ones drift. But I'd prefer just requiring them all to pass.
The generated CI files have Python lint steps. On GitHub, the lint steps will fail the pipeline if there are any errors. On GitLab, they only warn and the pipeline continues. The behavior is inconsistent. Our pure Python packages fail the pipeline on lint errors, so I think the GitHub rule would be preferable here. If some repos really need to skip linting, I'd rather have a way to opt out for the misbehaved ones than letting all the good ones drift. But I'd prefer just requiring them all to pass.