[imp] Questions with regard to CVS

[Nigel Cass]

Hi All.
In early September I hope to roll out a webmail system for our student
population. My initial tests with horde 2.1, imp 3.1 and turba 1.1 have
been very successful and I believe I am nearing launch capability :)

I do have a couple of questions with regard to updates via CVS though.
When I first set up my pilot system I just downloaded the latest release
tarballs from the horde ftp site. Since then I have read this list and I
notice that most users seems to frequently refresh using CVS.

I'm new to using cvs but I've installed it and tested it and I'm happy
that I know how to checkout the three modules that I need to use.
However I'm unsure as to how often I should be refreshing. What I really
want to do is to always be using stable versions (i.e preferably the
release tree.) but I'm concerned that my system then ends up too far
behind as it seems that config files are changing - for e.g some of the
config files I use in my version have now disappeared as far as I
understand. Also some have changed name I believe.

Is it just the case that when a new release version comes out I have to
grab it at that point and then re-configure it allowing for any changes
in the underlying configuration ? - do the <file>.php.dist files change
dramatically and if so do I need to copy my directives/config lines from
my configuration files into the new .dist versions every time I update ?

What about serious bugs that get fixed ? - do file versions instantly
get tagged as RELENG_x - in order that a "cvs co -r RELENG_x
horde/imp/turba" will download safe versions ?

Is it possible/advisable to download single files or should I always be
updating the full module ? 

What do you guys do with your live systems ?

TYIA

Nigel Cass
Systems Engineer
University of Hull