On Fri, Oct 11, 2013 at 09:28:38PM +0200, [email protected] wrote: > Even the human recipients sometimes have trouble to decide by content > what is spam, so a automatic detection for such a diffuse target is > doomed to fail.
This is plainly false. A filter does not have to detect all spam.
All it has to do is detect most spam, and rarely generate FPs.
There is no "doomed to fail" if the problem is correctly defined.
Anyway this is off topic. Can we stop?
--
Viktor.
