some SPAN results

[Tom Anderson]

On Mon, 2005-07-18 at 19:47, David Relson wrote:
> With a lexer modified to ignore <span...>...</span>, I scored the 6672
> spam I've received so far this month.  19 of them had different scores
> (numerically) depending on whether I used standard bogofilter 0.95.2 or
> the modified version.  None of the 19 classified differently.

Try piping them through Stripsearch
(http://orderamidchaos.com/bogofilter/stripsearch), and I'll bet all of
that "innocent" text doesn't offset a well-registered "SPAM-ADDRESS"
token.  Almost none of those long story or thesaurus spams get through
anymore.  I might have had one false negative this week.

Tom