Posted 3 years ago in
Git Rebase allows us to rewrite Git history. It is a common practice to use
git rebase to squash commits before creating or merging a pull request; nobody needs to see that you fixed 10 typos in 5 separate commits, and keeping that history is of no use. So how does a rebase look like?
Let's imagine you have your deck of cards, they are ordered in a certain way that cannot be changed. Each card represents a commit in a project's branch.
When running an interactive rebase with
rebase -i, there are mainly three actions we may want to perform in a commit (card):
- pick: pick a commit.
- squash: squash this commit into the previous one.
- drop: drop this commit altogether.
In this game, you want to squash cards together into doubles and triples. Some cards make sense on their own, so you will pick them. Sometimes, a card should not even be there, so you might want to drop it.
Although there are other ways of using
git rebase, the interactive rebase used like this is a common practice observed in projects that rely on multiple contributors, both in open as well as in closed source. It enables you to commit earlier and with more frequency, because you are able to edit your history before submitting your pull request.
If you'd like a deeper introduction to Git Rebase, please check this great dev article from @maxwell_dev: