On Wednesday 25 May 2005 17:34, Tom Collins wrote:
> Modifying the source to vdeliermail is also an option.  If, for
> example, you wanted to read the headers of the message and save it in a
> special mailbox for the domain if it's spam (or even save it in an IMAP
> mailbox for the user), it would best be accomplished in vdelivermail.

Forgive my opinion, but I think that this is the wrong (certainly not the 
best, anyways) approach.

Why not call your choice of LDA, such as maildrop or procmail, from either 
per-user .qmail files or .qmail-default?  That's going to be a much more 
flexible and easier to implement solution than hacking up vdelivermail.

This is what we do, for every user on our system.  The script we use to create 
accounts (which also is what's used by our web interface) sets up a .qmail 
file for the user, creates a number of folders we like them to have, and 
writes out a subscription file that the IMAP server uses.  We call clamav, 
dspam, and spamassassin from maildrop, and filter accordingly:  Spam to a 
filtered folder, and virii to /dev/null. :)

Users can then set up their own filtering rules as well, if you choose to 
allow it.

Casey Allen Shobe | http://casey.shobe.info
[EMAIL PROTECTED] | cell 425-443-4653
AIM & Yahoo:  SomeLinuxGuy | ICQ:  1494523
SeattleServer.com, Inc. | http://www.seattleserver.com

Reply via email to