I'm having a go with Subversion as a replacement for CVS at work. We've been pondering moving to it for a while, because we want something that's basically CVS but with the design mistakes taken out. That seems to be what Subversion is aiming for.
Subversion isn't in Debian stable, but there are backports. Fortunately for me they had already been set up by my colleage Aaron before I started looking in to this stuff.
The documentation for Subversion seems surprisingly good, although it took me quite a while to find the bit about how to create the repository. I may have just missed something. Anyway, this is the magic command:
svnadmin create directory
where directory is the directory you want it to create to put everything in. Easy.
I've managed to use the cvs2svn program to import the CVS
data for our website. Turns out that by default it does the repository
creation step above automatically, unless you tell it not to.
I'm particularly interested in using Subversion for our website, because it might allow us to rename or shuffle about files on the site without having to keep track of all the redirects necessary to keep old URLs alive. The plan is to have that done automatically using the history of file renames that Subversion provides. I have no idea yet how hard it will be to dig that information out, though.