Linus is dubious... but we get some praise

[Tom Anderson]

On Sun, 2004-01-25 at 10:53, David Relson wrote:
> Indeed, a lot of ideas and testing have gone into spambayes and there
> have been many interesting discussions.  Sometimes it surprises me that
> bogofilter, with its small core team, does so well!  Evidently we have
> enough of what's needed.

From: http://spambayes.sourceforge.net/background.html

"About the only "general rule" that kept showing up was _Stupid beats
smart_. That is, a fiddly and delicate piece of tokenization magic would
often produce worse results than something that just took a brute force
"just grab everything" approach."

I don't know whether that says anything about the average intelligence
of the bogofilter team ;)  But I think it just reinforces the idea that
adding a bunch of clever tokenization rules is the wrong method, and
that sticking to the core bayesian mathematics and simplified code is
the ticket.

Tom 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://www.bogofilter.org/pipermail/bogofilter/attachments/20040125/cdcd9b11/attachment.sig>