Phil Nelson <phil%NetBSD.org@localhost> writes: > On Friday 02 October 2015 09:56:02 Timo Buhrmester wrote: >> > ...because lying about history is good? >> Lying about history is bad, iff the history was written already. >> >> Nothing wrong with rewriting the present to become neatly organized >> history. > > This is one of my issues with git. Because someone else pushed > something before I pushed, with this workflow I now have to do a > rebase rather than just an update and a commit. I have personally > found that git rebase to be much more painful than the corresponding > CVS update and commit. This is not a fair complaint. The issues with git rebase in terms of merge conflicts with local commits are exactly the same issues with cvs or svn and updating with uncommited changes. It's just that in cvs/svn there is no choice. In all seriousness, if you have changed files that didn't change in the new commits, the rebase should require zero human intervention. When looking at history later, the merge commits from 'git pull' are just moise, since there was no intent to have a separate branch and merge. The reader is better served by a linearization.
Attachment:
pgpfJKi9ylzdF.pgp
Description: PGP signature