Hi All, I'm reading up on spamassassin and didn't realize that it was possible to train the Bayesian filter. Does anyone use spamassassin that has it configured so that users can report mis-identified messages? That's the one thing that I need to learn how to do before I can use it in production. c