... is the setup I've been using. In my configuration, I relegate Spamassassin to a couple of useful blacklists I wouldn't use with spamdyke. These are a bit less accurate so they need a more soft failure than spamdyke's go/no-go approach. What really adds to the game is the scan the body content to do some it's classification. This seems pretty effective as most spam that it high scores (10+) while good email usually get's close to 0.
I was just interested in what backend checks that others use with spamdyke, if any, and how they configure them. Since spamdyke kills most of the spam right up front, it seems that it significantly reduces the overhead for back-end processing. Gary _______________________________________________ spamdyke-users mailing list [email protected] http://www.spamdyke.org/mailman/listinfo/spamdyke-users
