Hallo, Ismael Valladolid Torres hat gesagt: // Ismael Valladolid Torres wrote:
> El martes, 9 de septiembre de 2003, a las 19:54, Frank Barknecht escribe: > > I run spamassiassin and regularily train its Bayes filter. > > Even though that would be pure off-topic, I'd love to know about > training spamassassin, as who trains a hamster. :) I added this to my mutt config: # Add this to your ~/.muttrc to integrate spamassassin and mutt # spamassassin report and classify as spam macro index S "|/usr/bin/sa-learn --spam --no-rebuild --single" # spamassassin classify as ham macro index H "|/usr/bin/sa-learn --ham --no-rebuild --single" # rebuild the spamassassin bayes filter db macro index R "!/usr/bin/sa-learn --rebuild" If a spam mail did get through spamassassin's own rules, I pipe it through the "sa-learn --spam" command to let spamassassin learn it, before I delete it. I also route some ham through the --ham line regularily. This traines spamassassin's Bayesian filter rules. For details see "man sa-learn". ciao -- Frank Barknecht _ ______footils.org__
