A new colleague of mine says he reads not the changes, but each commit, and will not approve if one commit wasn't good. Have you ever heard that before?

📌

well that's just silly :)

@netikras he was hired as a senior and seems to have a bit of an elitist attitude

Its a bit harsh but on the other hand, if you need to look at the history its the commit messages you see.

Both should be clear and relevant.

I more than once have found commit messages like “stuff” “to much to write” and such.

Those days I actually would side with your colleague ;)

But I guess he has a standard quite a bit above that. And from your rant, probably to high.

@Voxera I was under impression OP meant commit contents, not the message. If it's message we're talking about then I agree with the reviewer. But if we're talking about commit contents then I'm with OP.

@Voxera sums it up nicely once again

You guys don't just squash before merging? Only reason I see to keep an ancestor commit is if it contained some previously agreed upon design that you want to keep in the upstream for documentation purpose.

@netikras no, I totally agree with meaningless messages, they need to describe changes.

I am talking about the content of a commit. Say I made a commit removing something, then another adding the line again as I made a mistake, he would not approve it

@beegC0de that's the thing, we squash as part of our merge

@LombArd so... He's only reading one commit?

@beegC0de no it's not squashed prior to that, at the merge request it's comprised of multiple commits but squashed after approval and merging takes place (GitLab option)

@LombArd so just squash before pushing to your remote branch my man