Re: Problem with Author Label

[Larry Jones]

Rondal Ellifritt writes:
> 
> However, this is an incremental process, and there are still a few
> glitches. One is that, when code changes are committed, the 'author'
> label on the revision is always being set to me. I suspect this is in
> part a consequence of the fact that everyone is using the same unix
> login. (Yes, I know this is a really bad idea, and I'll eventually get
> people to change that behavior. But this is currently a constraint on
> the problem, so please don't try to "solve" it by telling me to change
> the constraints.) We're using different CVS logins, and we're using
> the 'cvs login' command with our own cvs id's, so I expected the
> 'author' label to be picked up from the cvs login id. (I know it's not
> being picked up from the unix login id, because the production login
> that we're using doesn't have a cvs login id.)

You seem to be suffering from a common misconception.  The CVS login
command just records the password for the current CVSROOT setting, it
does not have any effect on what user subsequent CVS operations will be
done as.  The user for a particular CVS command comes from the CVSROOT
setting for the affected directory.  The CVSROOT for a particular
directory is the one specified by the global -d option on the command
line, if any; failing that, it's the one recorded in the CVS/Root file,
if any; or, as a last resort, the one specified by the $CVSROOT
environment variable.

If you're all using the same Unix login, you're probably all using the
same working directory, which means it's going to be inconvenient to get
things to work the way you want -- you'll have to always specify CVSROOT
on the command line.

-Larry Jones

Like I'm going to get any sleep NOW. -- Calvin