I got 4-5 spams overnight that scored very low, and one didn't hit a single rule.
Hmm, the message looked familiar, and I checked my spam folder. Sure enough, there was one in there. The first 6 lines were identical to the one you posted, though the pithy saying, closing, and URL were different.
They all came through the Debian mailing list, so I assume a bunch of Bayes ham training allowed coming through that list to "help" the spammer? Any suggestions?
Mine came straight to me, so it tripped BAYES_99 easily. I've bumped up the score on the Bayes rules a bit, so that, RCVD_NUMERIC_HELO, and RCVD_BY_IP landed it a score of 5.3.
Depending on how consistent the rest of them are, it might be possible to track some of the phrases.
-- Kelson Vibber SpeedGate Communications <www.speed.net>
