Re: FreeBSD CVS Question (answered!)

> From: Max Laier <max_at_love2party.net>
>
> Make sure to have "cvsroot-all" in your sup-file as well, that should provide 
> all the needed information to get the ids right.

  You nailed it in one shot.  I missed cvsroot-all just having done the cvs
init and not even thinking to check for extra filesets.  Now the files check
out with the right expansions.  Thanks!

> From: Garance A Drosihn <drosih_at_rpi.edu>
>
> Sad to say, I suspect you did a lot of extra work that you could
> have avoided.

  Doing work, even needless work sometimes, is where I learn things.  (:

> It depends on what you mean by "a local CVS checkout".  On the
> machines which are doing a CVS checkout, how are they getting
> to your local copy of the CVS repository?  If that repository
> is NFS-mounted, then you are probably in good shape.  If your
> machines are doing a CVS checkout over RSH or SSH, then your
> server may be doing a LOT more work than you realize.

  Right now it is just on the same machine.  You're right that SSH would
probably involve extra overhead such as the encryption (at least for SSH).
I've tended to avoid NFS since the bad old days of RPC problems (<cough>
like 10 years ago), so perhaps I'll give that a try again.  (:

> But really, you'd probably be better off investigating the
> cvsup-mirror port.  It did not take me too long to setup,
> and it has worked very well for me.

  After reading and digging, it is like one supfile in a crontab plus the
CVS checkout.

> From: Tim Kientzle <kientzle_at_acm.org>
>
> Don't use RCS, use CVS.

  I am using CVS.  I figured that RCS did the $-header conversion on the
far side (under the control of CVS).  I was wrong -- apparently that was
under the control of something that I picked up when I included
"cvsroot-all".