On Sat, Aug 09, 2003 at 11:04:01PM +0000, Jonathan Hutchins wrote:
> X-Spam-Status: No, hits=-12.6 required=5.0
> 	tests=AWL,BAYES_10,EMAIL_ATTRIBUTION,IN_REP_TO,QUOTED_EMAIL_TEXT,
> 	      REFERENCES,REPLY_WITH_QUOTES,USER_AGENT
> 	autolearn=ham version=2.55
> X-Spam-Level: 
> X-Spam-Checker-Version: SpamAssassin 2.55 (1.174.2.19-2003-05-19-exp)

> I think that's the thing, I haven't been able to do the Bayesian training yet, 
> and I'm getting loads of crap to do it with.
> 
> How imporant is it to have a "ham" - which I assume means genuine messages - 
> database as opposed to just a spam collection?

Don't know.  The README says you need some.  I collected spam for a while and
fed it to sa-learn.  Then I fed it /var/spool/mail/jim as ham.  I usually have
200-300 messages hanging around in my mail spool.  I have "auto_learn 1" in
/etc/mail/spamassassin/local.cf so it is now learning from mail it is REALLY
sure about.  Notice your message had "hits=-12.6" and "autolearn=ham".  I don't
know what the threshold is for "REALLY sure", it seems to work so I don't pay
much attention.
-- 
Jim