On Fri, 28 Mar 2003, Professional Software Engineering wrote:
At 16:52 2003-03-28 -0700, Ross Simpson did say:
In researching this, the only answer I found was to move from a .forward
file to a .procmailrc file in each of the affected users' home
directory, and then put whatever rules I wanted to in there.
But, as I said, I don't want to do this. I'd like to have incoming mail
for that set of users be processed by procmail, then be forwarded on to
the fowarding address.
I don't follow why /etc/procmailrc isn't appropriate for your purposes:
I don't know how Ross' system is set up, but I'm using sendmail with
procmail as mda. In this setup, if a user has a .forward file with
simple email addresses, procmail is never called (/etc/procmailrc or
.procmailrc).
Now, as you suggested later in your email, I think you could add an
entry to /etc/alias for any such user and the alias would get
interpreted before the .forward, and give you a chance to stick
procmail in the middle. Of course, this solution doesn't seem
particularly scalable to me.
I've personally never found a "simple" way to do things, and I tell
users who want to get the benefits of our system wide filtering to use
a .procmailrc file that simply acts like a .forward file, meaning that
/etc/procmailrc is still read. In other words, I (or they) don't have
to add all the system-wide recipes to their .procmailrc, they just
have to have some .procmailrc so the system-wide recipes get called as
well. Considering that a .procmailrc acting as a forward is only
slightly longer than a .forward this seems most reasonable to me.
Nonetheless, if anyone knows of a way to force sendmail to invoke
procmail regardless of a .forward, I'm all ears (though it also seems
this would break the expected functionality of .forward thus confusing
more knowledgeable users).
--
Public key #7BBC68D9 at | Shane Williams
http://pgp.mit.edu/ | System Admin - UT iSchool
=----------------------------------+-------------------------------
All syllogisms contain three lines | shanew(_at_)shanew(_dot_)net
Therefore this is not a syllogism | www.ischool.utexas.edu/~shanew
_______________________________________________
procmail mailing list
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail