spamassassin doesn't seem to be using bayes

[jdow]

From: "D. D. Brierton" <darren at dzr-web.com>
> I'm using FC4 with spamassassin-3.0.4-1.fc4. fetchmail delivers mail to
> a locally running postfix. spamd is running as a service, and spamc is
> called by procmail on my mail. My setup is almost identical to that
> desribed here:
> 
> http://wiki.apache.org/spamassassin/UsedViaProcmail
> 
> However, despite the fact that I have trained spamassassin on a vast
> amount of both ham and spam using sa-learn, I suspect that Bayesian
> testing is not being applied. I became suspicious that this might be the
> case after receiving over a dozen almost identical messages and despite
> training spamassassin on them they are still not being identified as
> spam. So I started looking at the headers that spamassassin adds to each
> message more closely. Here is the header it added to a recent message
> from this list:

"Are you sure you trained on a vast amount of ham and spam?"

What is the basic mail format on your system? If you are using the
"mbox" format, the one which has a single file in /var/log/spool/<name>
rather than having a collection of files under a /var/log/spool/<name>
directory and you did not include "--mbox" in your salearn command
you get to start over. When training you should have seen a series of
periods, one for each message, and a note about messages seen and
messages used for training. If this is always "1" and "0" respectively
you have the mbox foulup.

{^_^}   Joanne, who has a VERY long screed about my arrogant opinion
        about how to care and feed SpamAssassin. (Turn off autolearn
        and auto-whitelist features, use (lots of) rules from the
        SpamAssassin Rules Emporium, and if at all possible enable per
        user scores, bayes, and rules are chief among my opinions. I get
        very good results.)