As I've been talking about, I've finished a first pass at hooking up
weighting into my "spamcheck.rc" file. The intent is to be able to
code things like "If From:=Reply-To: then this message is 30% likely
to be SPAM". To avoid any confusion that could result from adding
percentages, I've used procmail's scoring mechanism directly.
What was a single "spamcheck.rc" file is now split into several
smaller files which makes it more difficult to include in an email
message. Thus, I've set up a (perhaps temporary) FTP site at
ftp://ftp.bristol.nl/pub/users/DanS/spamcheck
Documentation of any kind is lacking, and I've only made quick guesses
at weights for various heuristics. I also haven't hooked up any
"positive" heuristics yet either such as "a match on ^FROM_DAEMON
(most mailing lists) is only 10% likely to be SPAM".
I'm especially interested in any comments you may have about the
design of "spamcheck", I'm trying to build the framework (using
primarily just procmail) for something that can expand and change as
needed.
Dan
------------------- message is author's opinion only ------------------
J. Daniel Smith <DanS(_at_)bristol(_dot_)com>
http://www.bristol.com/~DanS
Bristol Technology B.V. +31 33 450 50 50, ...51 (FAX)
Amersfoort, The Netherlands {info,jobs}(_at_)bristol(_dot_)com