From: Peter Pak <pakp(_at_)cs(_dot_)acu(_dot_)edu>
We are also looking for something to handle this problem, but we are
interested in having a bounce message go back as if from the
mailer-daemon with an appropriate error message. By the time the
$HOME/.procmailrc is read, procmail is running as that user so the recipe
probably needs to go in /etc/procmailrc. Has anyone done this? Is it
better to just throw mail into the bit bucket?
It would depend on whether the account is only going to be disabled
for a short period (e.g., a single term), or if it's a long period
of disabling followed by certain account deletion. Since many of
our students take a few quarters off and then return, we just toast
the mail and then remove the .forward upon re-enabling.
One student's account that we disabled (before the current scheme
was in place) received so much mail that after trying more friendly
methods, one of our sysops simply placed an X in front of the
username in the passwd database. Mail from the talkative mailing
lists bounced because "foobar" was no longer a valid user at our
site. The list servers -- being well-behaved and intelligent --
automatically unsubscribed him after two or three bounces (i.e., in
under twenty minutes).
One single mailing list owner, instead of unsubscribing him,
finger'd our site using foobar's last name; fingerd politely matched
the GECOS field, and suddenly Xfoobar was subscribed to the list, and
receiving $MANY Kbytes of mail. Our sysops were less than thrilled on
that occasion. Since they had seen me do Odd Things with procmail, I
was asked to come up with a scheme for toasting mass-mail for disabled
accounts. Thus was born the setup I showed earlier.
Just my two timeslices, and an example of how difficult it can be
dealing with mass mail. We eventually decided that a near-blanket
solution was better, and if the student comes back within a year,
all of his personal mail will be waiting, and his mailing-list stuff
will magically start arriving again.
Luck++;
Phil
--
#include<std/disclaimer.h> The gods do not protect fools. Fools
finger pedwards(_at_)gamma(_dot_)cs(_dot_)wright(_dot_)edu are protected
by more capable fools.
email pedwards(_at_)valhalla(_dot_)cs(_dot_)wright(_dot_)edu
-Larry Niven