Problem: Whether the coverage/diff tooling correctly treats a moved file as a rename (versus a delete-plus-add, which would wrongly zero out its prior coverage history and could trip the patch-coverage gate on an otherwise-unchanged file) is unverified today — including for the already in-flight queue decomposition work.
Area: ORB / CI tooling
Proposal: Do a small, throwaway test move of a real file and confirm how Codecov and the diff tooling represent it, before any real extraction work in this milestone begins.
Deliverables:
- A documented, verified answer: does a plain
git mv read as a rename to the coverage tooling, or does it need to be structured differently (e.g. moved in a commit with no other changes) to be recognized correctly?
Acceptance criteria:
Boundaries:
- This is a one-time verification, not itself a real code move — revert the throwaway test move once the answer is confirmed.
Problem: Whether the coverage/diff tooling correctly treats a moved file as a rename (versus a delete-plus-add, which would wrongly zero out its prior coverage history and could trip the patch-coverage gate on an otherwise-unchanged file) is unverified today — including for the already in-flight queue decomposition work.
Area: ORB / CI tooling
Proposal: Do a small, throwaway test move of a real file and confirm how Codecov and the diff tooling represent it, before any real extraction work in this milestone begins.
Deliverables:
git mvread as a rename to the coverage tooling, or does it need to be structured differently (e.g. moved in a commit with no other changes) to be recognized correctly?Acceptance criteria:
Boundaries: