[Spambayes] I am the author of my own undoing.

[Bill Yerazunis]

   From: Webb Scales <scales at zko.dec.com>


   So, I'm pretty much in love with the procmail filter.  (And, I'm
   getting ready to have it just delete spam on contact, so that I
   *never* see it!  I'm really liking SB.)

   The obvious problem is, how do I training?  There are at least
   three answers.

Let me offer a fourth option:

Assuming SpamBayes has a way to command learning from the command line,
create a pair of "magic words" that nobody is likely to ever guess.

Then, change your procmail recipe such that:

 Whenever procmail sees magic word A, it invokes Spambayes in the "learn
 this as spam" mode.

 Whenever procmail sees magic word B, it invokes Spambayes in the "learn
 this as NONspam" mode.

Then, to train, you just include the magic word, and re-mail the message
to yourself.

Now, there are certain details that I'm glossing over (like messages
that get altered in your pipeline, headers lost and added, etc) but this
is a good starting point.

I've been using this technique on CRM114 since inception, and it 
seems to be quite easy, as all it requires is a working mail-sender
that you can type and forward with.

     -Bill Yerazunis