How does this look for a way to "un-poison" SA-Talk:
It uses the Procmail "OR" scheme of checking for the failure of ANDed
negation.
--- procmailrc ---
:0 c
* ! ^X-Spam-Status:.*autolearn=ham
* ! ^TO_\/[EMAIL PROTECTED]
* ! ^TO_\/postfix-users@(postfix.org|cloud9.net)
{ } # This evaluates to an Or statement
:0 Ec
| /usr/bin/sa-learn --forget
-------------------
Not sure if I need to clone,
<<Dan>>
| -----Original Message-----
| From: Matt Kettler [mailto:[EMAIL PROTECTED]
| Sent: Thursday, December 04, 2003 10:22 AM
| To: Smart,Dan; [EMAIL PROTECTED]
| Subject: Re: [SAtalk] SA-Talk poisoning Bayes
|
| At 11:06 AM 12/4/2003, Smart,Dan wrote:
| >In reading the sa-learn man file, it says running
| discussions of spam
| >through sa-learn is bad. Does SA take this into account already, or
| >should I create a procmail rule to bypass SA for messages
| from SATalk
| >and
| >(possibly) Postfix-List ?
|
| SA's bayesian system does not take into account where email
| comes from.
|
| So, you might want to consider creating a procmail bypass if
| you're concerned about poison.
|
| That said, I do run sa-talk through spamassassin here, and
| haven't had too much trouble, however my autolearn thresholds
| are set a bit wider apart than most.
|
| I do avoid sa-talk, razor-users, sa-dev, etc when setting up
| my training however.
|
-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills. Sign up for IBM's
Free Linux Tutorials. Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
Spamassassin-talk mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
