3 ms·
I follow your flow, and I don't bother squashing my commits. So the repo history has lots of dumb little commits. We use a branch model for features, though,
by rootusrootus 5y ago
I follow your flow, and I don't bother squashing my commits. So the repo history has lots of dumb little commits. We use a branch model for features, though, so I can just filter for those merge commits if I want to see granularity at the feature level.
- muxator 5y agoEasier for you. Harder for who reads you history.
- mkirsche 5y agoIn addition it makes rebasing harder, which might make you use merge commits and then reading your commits gets even worse because changes might hide inside those merge commits.
- rootusrootus 5y agoProbably. Like I said, though, it's easy to just look at the merge commits and ignore the rest. I don't personally read per-commit history almost ever. If I have to follow someone's train of thought that closely to understand what happened, then 1) the docs suck, 2) or there should have been comments and they are missing, or 3) the story was too complicated. I will stipulate in advance that this is very much going to vary by project, language, and experience. It is entirely legitimate to have a different point of view ;-).