On Tue, Jun 23, 2009 at 14:32, Henrik K<h...@hege.li> wrote: > On Tue, Jun 23, 2009 at 03:24:03PM +0200, Benny Pedersen wrote: >> >> On Tue, June 23, 2009 10:04, Henrik K wrote: >> >> > Thanks for testing. If you have bad memory, you might as well disable it >> > already. :) >> >> thanks for the code of this plugin and dns data, to know more what will >> happend i like to ask: >> >> will you update this plugin again ? >> will some results from the testing be shown in public ? > > Well, people posted their results.. also: > > http://ruleqa.spamassassin.org/20090620-r786769-n/T_EMAILBL_TEST_LEM/detail > > But most SA corpus include easy-to-block botnet-traffic, so freemail spam > looks like it doesn't exist..
that's ok, we take that into account. it does pretty well against zmi's corpus in particular. I suspect it'd do great with additional meta combinations. (reminds me. I meant to add a new view that only measures against spams that score < N where N== 8 or so, to avoid that.) --j.
