If you're on a feature branch, you can probably commit often, and push often (effectively backing up your work), then, the isolated unit of work is done on a si

If you're on a feature branch, you can probably commit often, and push often (effectively backing up your work), then, the isolated unit of work is done on a single commit (non-FF) when merging with the main branch (this is the one you search for when you want to revert something or find the culprit for something). This allows having "best of both worlds", I think. EDIT: Take a look at the feature branches on the git flow model: http://nvie.com/posts/a-successful-git-branching-model/ http://nvie.com/posts/a-successful-git-branching-model/