Hyperbolic trees would probably be the only way to visualize a largeish
cell without losing detail or going insane... I'm not sure what CVS
really has to offer Coda installations except maybe something ubiquitous
to sit on /coda.  eg. instead of using pserver most of the time, define
/coda/cvs as CVSROOT on all clients and let Coda take care of keeping
things up to date; only use pserver under ssh to update through the
firewall.  

Where I work we now check damn near everything that we give a hoot about
into CVS, and I back that up each night.  Sooner or later I will start
replicating the CVS tree (though I have to figure out how to do this in
a diff-and-patch way) onto multiple servers.  If I can't think of
something better, I may just try backing up off of the Coda volume with
a throwaway CVS tree on /coda, and when I'm satisfied that the data is
solid, start using /coda/cvs for our main CVSROOT.

We plan on doing numerous stupid and amusing things with/to our Coda
servers when I have more time; at present I have a big project about to
be rolled out, so nothing new or interesting will happen here with Coda
for a week or so.

-- 

    "Of course I'm crazy, but that doesn't mean I'm wrong."

                                      --Anders Hammarquist