Re: [math-fun] Spam-blocking, Bayesian methods