Re: Bayes/spam Issue

[lvlist]

Thanks for your help.  I have been reading and it keeps asking for hams and 
spams.  In my setup I am unsure if I am sending copies of my spam to a 
directory on my box.  I am sending a copy of every spam to a mail box on my 
main email server.  This is my relay server running postifix and 
spamassassin out in the DMZ.  What should I change in the spamass. config 
to send a copy to a directory.  I am really confused on how to seperate 
hams from spams. I will keep reading.  Any pointers would be great.  Thanks 
again.

Les
At 04:32 PM 3/21/2004, you wrote:
http://wiki.spamassassin.org/ look up training. There are too many
different setups for me to expound at length. What I happen to know
moderately well is fetchmail, sendmail, and procmail mushed together
to make it all work. Initial training involved collecting spam in
one place and ham in another, both in "mbox" format. Then feed it to
sa-learn something like: sa-learn --ham --showdots --mbox ~/train/ham
That wants to change modulo your setup.
{^_^}
----- Original Message -----
From: "lvlist" <[EMAIL PROTECTED]>
To: "jdow" <[EMAIL PROTECTED]>
Sent: Sunday, 2004 March, 21 03:51
Subject: Re: Bayes/spam Issue
> ok, but I have not idea how to go about this.  Can you point me to a more
> specific reference?
>
> Thanks for your response.
>
> Les
>
> At 03:45 PM 3/20/2004, you wrote:
> >From: "thelvlist" <[EMAIL PROTECTED]>
> >
> >I am new to spamassassin and I am running it on postifix as a mta relay.
It
> >seems to work ok but I am getting lots of spam now.
> >
> >I am using: SpamAssassin version 2.54 (Also very curious about the best
way
> >to upgrade.  I had a hell of a time getting it to work the firs time)
> >
> >I am trying to figure out if spamassassin is actually working right and
> >using bayes.  I looked at all the faq's and I have this info:
> >spamassassin --lint -D:
> >
> >[EMAIL PROTECTED] root]# spamassassin --lint -D
> >...
> >debug: Initialising learner
> >debug: debug: Only 0 spam(s) in Bayes DB < 200
> >^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> >
> >You are using Bayes in that it is in your configuration. You are not
> >using Bayes because it is untrained. Once it is trained it will become
> >active automatically. Read up on training Bayes. 200 hams and 200 spams
> >become your friend when you feed them, separately, to Bayes training.
> >
> >{^_^}
>