John Stewart wrote on Wed, 28 May 2003 16:09:05 -0500: > You want the headers! Bayes uses these.
That's correct, but if you look at a spamassassin -D output you see that the header information usually plays a marginal role. From a typical message you may have 1 - 3 matches in the headers but about 10 or more in the body (no matter if spam or ham tokens). So, the next statement isn't correct in my opinion: Forwarding will lose virtually all > the useful information. > However, you are definitely watering down the results if you don't have the original headers. Also, some Received lines will always be the same whether it's spam or ham, because it all goes thru the same mail server(s). Therefore I would like to see an option to tell SA/sa-learn just to use tokens from the body and leave the header alone. There are nice header tests already. Just using the body would ease it manyfold for normal mailbox users to submit spam/ham for learning. Kai -- Kai Schätzl, Berlin, Germany Get your web at Conactive Internet Services: http://www.conactive.com IE-Center: http://ie5.de & http://msie.winware.org ------------------------------------------------------- This SF.net email is sponsored by: eBay Get office equipment for less on eBay! http://adfarm.mediaplex.com/ad/ck/711-11697-6916-5 _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
