>-----Original Message----- >From: Ryan Thompson [mailto:[EMAIL PROTECTED] >Sent: Monday, August 02, 2004 9:57 PM >To: Pierre Thomson >Cc: [EMAIL PROTECTED] >Subject: Re: sample scoring zero > > >Pierre Thomson wrote to [EMAIL PROTECTED]: > >> Here's a sample with only a Bayes hit, and not very high at that... >> anyone get better results? >> >> http://frodo.bruderhof.com/norules.txt >> >> pt > >It scored 37.1 for me, but the SURBLs have picked it up since, and I >have some rather helpful rules of my own that pushed it well into >autolearn before SURBL even had a chance to fire. SURBL, of course, >still rocks. :-) > >Rule writing is addictive. Anyone else suffering from this debilitating >addiction, HELP IS ONLY A .cf FILE AWAY. Join RuleAnon: The easy 12.0 >Point Program to autolearn a better way. >
I fear it is far too late for many a ninja. We are beyond help :) "Its not a habit. Its cool. I feel alive." <- for Kai ;) And yes I was WAAAAY behind in updates this weekend. Thanks to all who helped me catch up. Also a note on this recent "Spam Bombing" technique. I notice these runs all use the same dir name. This run used /f74/. So I'm continplating a sort of quick rule update to the 70_sare_specific.cf file. And simply making a quick rule to look for these dir names. But it would need to be update more frequent, and especially on weekends. So this is still just a theory. But keep an eye on those dir names, they stay the same for the run. --Chris
