I'm at the stage now in a small project of my own where I've been using 
CVS internally for development - I'm familiar with creating new 
repositories, adding, commiting and merging within the one source tree 
and I haven't had many problems with it.

Now I'd like to make a release 'branch' and then make it publically 
available but so far I haven't had any experience with doing branches 
with CVS apart from seeing what the NetBSD team do on a regular basis.

I'm expecting that there will need to be some bugfixes along the line 
to the code as released and I also plan on going ahead with development 
with other features etc, so am thinking I will definately need at least 
two distinct branches of the source to work from ...

This seems to me something very similar to the NetBSD manner of doing 
things and I was wondering whether there were any notes about how the 
NetBSD core team uses CVS to manage releases, bugfixes and pullups and 
all that sort of thing ... so far I've not had much luck finding 
documentation on the web on CVS that gets beyond basic stuff like 
creating repositories, adding/removing files, commits, merging files 
and tagging.

I've done some playing with CVS to see if I could figure it out myself 
and it seems that tagging is part of the solution, but I'm not sure how 
to get CVS to treat two code branches independantly - any revisions I 
make to one get merged into the other whenever I do an update to see 
whether it worked ... most frustrating ...

Also, I'm well aware of the current 'CVS is dead, use 
{subversion|arch|Perforce}' thing, but I'd like to make up my own mind 
- and that means first working out how it works with CVS ... :)

-- 
Malcolm Herbert
Computer Support Officer
School of Geosciences
Monash University
ph 9905 4881