I just converted my cvs repository to subversion.
It is already much cooler than cvs.
I think I'll take a look at seeing what would be involved with making subversion use postgresql as a backend database (instead of, or in addition to, BerkeleyDB 4).
If all WebDAV shares on Mac OS X can do the iDisk local cache thing, I guess that that wouldn't be necessary (since you could connect to a svn webdav share and work disconnected), but I'm not sure if that's an idisk only thing.
If that's not possible, then using postgresql as the backend would work (with postgresql replication, if that works the way it should).
In any event, it will be fun playing with it.
