On Wed, 04 May 2011 18:02:18 +0200, Ryan Leavengood <leavengood@xxxxxxxxx> wrote:
On Wed, May 4, 2011 at 11:50 AM, Oliver Tappe <zooey@xxxxxxxxxxxxxxx> wrote:If someone has an idea how to create consistent revision IDs in Mercurial, please speak up, as otherwise, IMO, Mercurial won't do, no matter the vote.Even though I'm in favor of Git, the majority at this point seem to prefer Mercurial. I wonder if we could just drop the global revision ID concept and just deal with the SHA1 hashes for commits. Of course I feel the same on this regardless of whether we use Git or Mercurial. We are changing tools and may have to change a few other aspects of how we work. Saying Haiku revision 4d5b6789 is not as pretty as Haiku r43215 but it isn't unworkable.
When talking about consistent revision IDs in Git, are these the same as what is output by 'git describe' (tag name + number of commits since then + hash)?
Also I would be curious why so many people prefer Mercurial. My vote is pretty much just laziness: I know Git. From what I've seen Mercurial is pretty similar so I'm not going to be too upset if it wins the vote. But I'm curious if the votes are more like mine (people just already know Mercurial) or if there was a definite preference for Mercurial after using both. If the latter, why?
Git seems to be the best candidate to run on Haiku (no hard links or Python requirements).
Also, Hg seems to have a lot of (git's) functionality added on later by means of extensions (stash, rebase). That makes me wonder how well all that is integrated (but it could be fine, of course).
Also, if Tk ever gets ported to Haiku, we'll always have a (although ugly) up-to-date GUI support tools available (git gui and gitk).