I think it's time for v2 now before we begin our core raylib refactors (e.g. #36642).
Ultimately, v2 will record the shortest screen-recording possible to achieve 100% coverage over the UI, and it will discover that path on its own. On each UI PR, it will post the full recording from master and PR, each contiguous chunk that's different, and a nice composite diff views of those chunks.
There's a simple version of this that should be quick to get up and running and will be more valuable than v1:
- script the path/actions
- assert ~80% coverage, post the coverage report with the PR
- just do this for mici ui
I think it's time for v2 now before we begin our core raylib refactors (e.g. #36642).
Ultimately, v2 will record the shortest screen-recording possible to achieve 100% coverage over the UI, and it will discover that path on its own. On each UI PR, it will post the full recording from master and PR, each contiguous chunk that's different, and a nice composite diff views of those chunks.
There's a simple version of this that should be quick to get up and running and will be more valuable than v1: