Hi, > git pull -r*) URL says: [...] > branch.master.remote = <nickname> > and after cut-and-pasting that into .git/config, replacing > <foo> with values, I get [...] > fatal: bad config file line 17 in .git/config > where line 17 is [without any indentation] > branch.master.remote = eddy The confusion is that while you can set values with this syntax if you're using git config (git config branch.master.remote <some value>), they're written differently in actual config files: [branch "master"] remote = <nickname> merge = <remote-ref> [remote "<nickname>"] url = <url> fetch = <refspec> I can't think of a message of comparable length that would make this clearer, though. > Jan. -- who just finds out the uninstall target is missing?!? > See attached. I have no comments about the patch contents themselves, but please read Documentation/SubmittingPatches. Most importantly, patches are usually sent inline here, and without a signoff line they can't be accepted for inclusion in the official repository. > *) I've been wondering why GIT lacks a "update", ie > pull-without-merge command. You almost never want > to git pull /without/ -r? -r stands for "rebase", i.e. it makes git pull use rebase instead of merge. Rebase has certain problematic implications for decentralized development, which is why it should only really be used in specific circumstances (for rebasing commits that haven't been published in any way yet). "You" may almost never want to use git pull without rebase, but "many others" do, and very often. Personally, I donn't recall a single occasion where I have used git pull --rebase (not counting git-svn stuff). Anyway, look at config options branch.autosetuprebase and branch.<nickname>.rebase to see how you can make pull use rebase by default. -Jan -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html