Re: bayes before spamassassin?

[Justin Mason <jm...@jmason.org>]

Matt Kettler writes:
> jeaspam@yahoo.es wrote:
> > Hi List
> > It is possible that each incoming mail goes first to bayes and later to 
> > spamassassin?
> >
> > Do you think that this procedure is good?
> >
> >                                      __________
> > (Incoming mail) ------> |   (bayes)      |(yes)---->(delete it)
> >                                      | it's spam?    |
> >                                      |__________|
> >                                      (no, goes to)
> >                                             |
> >                                             |
> >                                             |
> >                                        (spamassassin)----->(move the message 
> > to folder /mailbox_user/.spam )
> >
> > I've attached the textart anyway if you can't see it
> >
> > How do I do it?
> 
> You could do that, but you'd have to use a separate bayes-only spam
> tool. At that point, why bother? SA has it's own bayes system that it
> will run when the message is fed to SA.

in SpamAssassin 3.2.0, you'll be able to do it with SpamAssassin's
bayes -- short-circuit to "spam" with a high score if BAYES_99 hits,
for example.

--j.