[Evolution] dspam


I'm actually a SpamAssassin (server-side amavisd-new/SA smtp proxy on
Postfix 2.1) fan, but it has immense drawbacks (Perl daemon, takes
enormous resources - memory, disk I/O - limited user choice without
procmail/maildrop filtering and others); so after seeing a couple of
glowing dspam reports on the Exim list I installed server-side dspam
(2.10.6) in the weekend.

I'm thrilled with it - it's pure Bayes based, but 100% individual
choice, works well with Evo (had to make an own mini-sh-script for my
Evo filter), trains easily, is fast to learn (after feeding it 1,000+
spam/ham corpi) and dead accurate after a couple of days' teaching. It
thought that all of Michael's posts on this list were spam to begin with
- strike me, I don't know why - but doesn't any longer :) Drawbacks seem
to be high disk I/O and rapidly growing recipient db4 Bayes databases.
But zilch memory overhead, no Perl (ANSI C).

Has anyone any experience with server-side dspam and Evo over an
extended period? Opinions?



We make out of the quarrel with others rhetoric
but out of the quarrel with ourselves, poetry.

mail: billy - at - billy.demon.nl

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]