[Spambayes] Randomized Spam Beating SpamBayes

[SpamBayes.20.firefairy at xoxy.net]

Is anyone else getting this?

SpamBayes catches all the "identifiable" spam right off, only a little bit
of training required, but there is a variety that makes it through almost
every time.  It consists of either an entirely random set of content (no
idea what the point to those are) or a small chunk of content followed by
lots and lots of drek.  The subjects are sometimes coherent, sometimes
gobbledygook.

Is there any way to train for random gibberish, or have they actually beaten
SpamBayes?

Quinn
~~~~~~~~~~
Machine-independent: Does not run on any existing machine.
~~~~~~~~~~