Larry,
Thanks heaps..this means..my CVSROOT/passwd file format would look like this?
<username?:<encypted password>:<username which cvs runs as>
Aditya
[EMAIL PROTECTED] writes:
>
> <unixprompt>$cvs status -v readme.txt
> setgid failed: Not owner
> cvs status: authorization failed: server <servername> rejected access
If you're going to run pserver as a non-root user, you *must* use a
CVSROOT/passwd file to map all of the cvs users to the user you're
running pserver as.
> I think i dont have the latest version of the server.c file (as I had to
> manually make the change on line 5084 instead of 5114 in your patch).
> How do I know what revision of server.c i am using? How do i get the latest
cvs
> tarball with any patches? I've got cvs 1.10.8
The patch was against the current development version, but `patch'
should have been able to apply it to 1.10.8 (with an offset). There
isn't any way to tell what version of a particular file you've got
(unless you've got it checked out from the public CVS repository), but
it presumably matches the CVS release you've got. There haven't been
any releases since 1.10.8, so the only way to get anything more recent
is to check out the current development release from the CVS repository
(details are at http://www.cyclic.com.), but you shouldn't need to do
that.
> P.S. -> Making sure i've got it right...
> on AIX, the initgroups(3) system call, prints a message on stderr and return
an
> error when the underlying call to setgroups fails.
Exactly. There's no reason to print a message, it should just return
the error.
-Larry Jones
He just doesn't want to face up to the fact that I'll be
the life of every party. -- Calvin
