On 03/21/03 03:14 PM, Tony Earnshaw sat at the `puter and typed: > fre, 2003-03-21 kl. 12:38 skrev Tom Pfeifer: > > > today I got a spam which obviously carefully designed fake headers > > in order to get a lot of negative scoring to compensate > > the spam scores: > > > > fake references --> IN_REP_TO -3.300 > > fake pine headers --> USER_AGENT_PINE -5.801 > > fake Mozilla/Linux -> USER_AGENT_MOZILLA_UA -5.801 > > > > So, maybe at least contradicting user agents should count > > once only, or be excluded at all. > > We had this this week, it was in a MACK Truck thread. Dieter Weber: > www.hughes-family.org/bugzilla/show_bug.cgi?id=1647 > > Macro for multiple UAs to put in local.cf, make the scores whatever you > like. Mind the line breaks, this posting is "folded" at 76 chars. > > With regard to fake references --> IN_REP_TO -3.300: Alter the score > in local.cf if you don't like it. Don't forget to restart spamd :-) > > meta TOO_MANY_UA ((USER_AGENT_MOZILLA_UA + > USER_AGENT_PINE + USER_AGENT_KMAIL + USER_AGENT_VM + USER_AGENT_MACOE + > USER_AGENT_GNUS_XM) > 1) > describe TOO_MANY_UA Doesn't make sense to have more than one > UA > score TOO_MANY_UA 18.0 > > meta TOO_MANY_UA2 ((USER_AGENT_MOZILLA_UA + > USER_AGENT_PINE + USER_AGENT_KMAIL + USER_AGENT_VM + USER_AGENT_MACOE + > USER_AGENT_GNUS_XM) > 2) > describe TOO_MANY_UA2 Way too many user agents > score TOO_MANY_UA2 40.0 > > meta TOO_MANY_UA3 (TOO_MANY_UA && REFERENCES) > describe TOO_MANY_UA3 Lots of user agents with references > score TOO_MANY_UA3 12.0 > > meta TOO_MANY_UA4 (TOO_MANY_UA && IN_REP_TO) > describe TOO_MANY_UA4 Lots of user agents with a "reply to" > score TOO_MANY_UA4 12.0 > > Best, > > Tony
Ok then, before these sneaky jerks start hitting me, I want to get this done. So I have a couple questions. These 4 rules just go into the local.cf file and that's it? What if I'm running spamassassin as myself (not as root or MDA user)? Thanks folks! BTW, I just upgraded from 2.44 to SA 2.50 and I REALLY am impressed. I've had much more extreme scores in general. Spam typically scores high (above 10) and non-spam typically scores low (usually well below 1). And the bayes scoring hasn't even kicked in yet! Nice job! Lou -- Louis LeBlanc [EMAIL PROTECTED] Fully Funded Hobbyist, KeySlapper Extrordinaire :) http://www.keyslapper.org ԿԬ White dwarf seeks red giant for binary relationship. ------------------------------------------------------- This SF.net email is sponsored by:Crypto Challenge is now open! Get cracking and register here for some mind boggling fun and the chance of winning an Apple iPod: http://ads.sourceforge.net/cgi-bin/redirect.pl?thaw0031en _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
