I would like to add a new section called "CI infrastructure overview" to the root CONTRIBUTING.md. This thing cannot only exist in my/Ralf's/Marcus's heads 馃檪
To the bare minimum, we need to add a diagram for visually explaining
- what are the artifacts that we generate in the CI?
- which runners are used at each stage (self-hosted vs GitHub-hosted)?
- how are artifacts stored and propagated in different stages of the CI pipelines?
- how are artifacts flowing across pipelines for different branches (main vs backport)?
- how are we utilizing various free (?) storage spaces (GitHub Artifacts vs GitHub Cache)?
It is fine to add a short writeup, but it should be succinct, accurate, and easy to maintain/update. I don't care if there is no writeup accompanying the diagram.
Since we are touching the CONTRIBUTING.md, we should also add a new component cuda-pathfinder to the list, and a footnote to cuda-bindings explaining that cuda-python also follows its contributing guide
I would like to add a new section called "CI infrastructure overview" to the root
CONTRIBUTING.md. This thing cannot only exist in my/Ralf's/Marcus's heads 馃檪To the bare minimum, we need to add a diagram for visually explaining
It is fine to add a short writeup, but it should be succinct, accurate, and easy to maintain/update. I don't care if there is no writeup accompanying the diagram.
Since we are touching the CONTRIBUTING.md, we should also add a new component
cuda-pathfinderto the list, and a footnote tocuda-bindingsexplaining thatcuda-pythonalso follows its contributing guide