On Thursday 23 February 2006 16:36, Jon D. Slater wrote:

...

I've been using Mailman for some time, and just recently installed SpamAssassin on my Fedora Core 4, Linux machine.

Spam-assassin would normally be used with the MTA (i.e. exim, qmail, etc.) and not Mailman. Basically, you should be using it to stop spam from getting to Mailman alltogther and not after the message has already been accepted by your SMTP daemon.

So this is likely not the place to be asking that question, try looking through the Spam-Assassin site for intergration how-tos for your SMTP daemon.

Oh, I disagree, it's not that bad a place to ask - How about Mailman FAQ #4.23? http://www.python.org/cgi-bin/faqw-mm.py?req=all#4.23

Three methods -

#1 is MTA integration

#2 uses patches - This is what I do, and it works great. Unfortunately, the link http://jamesh.id.au is no longer good, so you will have to get the patches elsewhere. Google spamassassin.py or see http://sourceforge.net/tracker/index.php?func=detail&aid=640518&group_id=103&atid=300103 You will need to add the following to your mm_cfg.py: # GLOBAL_PIPELINE.insert(1, 'SpamAssassin') SPAMASSASSIN_HOST = 'localhost' SPAMASSASSIN_DISCARD_SCORE = 500 SPAMASSASSIN_HOLD_SCORE = 3 SPAMASSASSIN_MEMBER_BONUS = 0 # You can also do a similar thing for Clam if you get that patch # Scan messages for viruses using Clam AntiVirus GLOBAL_PIPELINE.insert(1, 'ClamScan') CLAMAV_DISCARD = 0 CLAMAV_CLAMDSCANPATH = '/usr/bin/clamdscan'

#3 Procmail also works, of course.

John