I defend atomic PRs and squashing merge commits. It's simple, makes reviewing easier and works for 90% of the cases. Maybe I'm missing something, but I find

I defend atomic PRs and squashing merge commits. It's simple, makes reviewing easier and works for 90% of the cases. Maybe I'm missing something, but I find it painful to rebase into a neat history, and my colleagues would probably hate me to police them around to do it as well.