I’m determined to beat Flow to Github!
I’d like to point out that it’s not my fault that Smack isn’t using git: http://community.igniterealtime.org/message/227448
Once the repo is moved to under the igniterealtime organization, I will setup a cron on one of our servers over here to sync the SVN and Git repo’s together, using the SVN as the master for now. I’ll set it to run at GMT midnight since we have a pretty international community, hopefully that is convinient for everyone.
Mirroring svn <-> git is usually not an good idea. It just adds unneeded addtional complexity. The best approach is to set the svn repo read-only and start using git. But I am not a Spark developer, that this is not really something that concerns me.
I figure we will have some transition period where we are still using both repos.
I don’t see a reason why we should have such a transition period. It only allows people to delay the unavaoidable, which will only cause us more work in the end.
Git repo structures are usually slightly different than the normal SVN repo
Correct, but the changes are trival. The biggest difference is that git doesn’t keep tags and branches in special directories. Instead, branches and tags are pointers to commit SHA1s. Simply as that.
For every new (or experienced) git user, I really recommend reading “Pro Git” by Scott Chacon, which is freely available under http://git-scm.com/book
Oh and by the way, before everybody get’s to much hyped by git: Robins answer to http://community.igniterealtime.org/thread/51292 is still missing.