I've been putting sa-learn through the gears with many thousands of spam messages (gotta love web hosting 100+ domains most of which do nothing but collect spam /sigh).
I'm curious how Bayes is *supposed* to be learning... I find that despite learning from hundreds of MB of spam that spam is still getting through.
Looking through my headers and the rules, etc., should I not be seeing more BAYES_## items in the headers to see that the Bayes is, in fact, learning what I'm calling spam and treating it according to add points?
Thanks, ian
Yes, you should be seeing bayes hits, which leads me to belive you have a configuration problem.
First, run spamassassin -tD <sample-spam.txt.. look at the debug output. Is bayes even enabled? are there enough tokens?
How are you calling SA (procmail, mailscanner, amavis, etc etc)? What user does SA run as when you get email? What user did you run your training as?
------------------------------------------------------- This SF.Net email sponsored by: Free pre-built ASP.NET sites including Data Reports, E-commerce, Portals, and Forums are available now. Download today and enter to win an XBOX or Visual Studio .NET. http://aspnet.click-url.com/go/psa00100003ave/direct;at.aspnet_072303_01/01 _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk