On Wed, Jun 18, 2025 at 02:51:02PM +0200, Marc Haber wrote:
How about force pushes after doing rebase -i to squash fixups? I have been told to NEVER do that, but the one time I accidentally did it, it just worked.
This was the blanket advice people were giving 15 years ago, but nowadays most people take a more nuanced view and are fine with rewriting history in work-in-progress branches (which most MRs correspond to). I routinely force-push to in-flight MRs. Nowadays GitLab even retains history across force-pushes so that reviewers can still go back and look at earlier versions if they need to.
The way I look at it is that the sequence of commits with their logical separation of changes and their associated commit messages is part of how you communicate the intent of the change you're trying to make, and adjusting that sequence of commits is a typical part of how people should respond to review.
-- Colin Watson (he/him) [cjwat...@debian.org]
