RFC: New criteria for commit access

This is being addressed:

Those and previous threads are trying to do precisely that, but there doesn’t seem to be consensus on that one.

I’m a strong believer of code review and I don’t understand how a project as big and important as LLVM can still be pushing for merge without approval.

But that, in itself, won’t close any holes. It’ll only increase the cost of supply chain attack to two new people approving each other’s patches.

Making it easy to get commit access (including the need for stacked PRs) means that’s still very cheap to do.

2 Likes