[Gnu-arch-users] cvs interoperation question

[Jani Monoses]

Hello

what I want is track a CVS upstream where I have commit access and
provide a local arch hub for myself and other devels who don't.
in the wiki at
http://wiki.gnuarch.org/moin.cgi/Arch_20and_20CVS_20in_20the_20same_20tree
it says that one should unprecious CVS files so they are archived too.
What benefit does that have if commits/updates to/from CVS only happen
from one working dir? The same doc further on says one should commit
from a single place to avoid complications.

Does the setup below look ok? What do others use in a similar situation?

ex:
Having project--cvs--1.0 the main branch with only one tree checked out
and which is a mixed CVS/arch working directory. This is where cvs up is
done periodically and also form where I commit.
Other branches tag from main and can have more project trees checked out
and they are arch only project trees w/o CVS metadata.
>From time to time changes are merged into the main tree from other trees
and committed to CVS so arch main and the CVS repo are always just one
update or commit from eachother.


thanks
Jani