On Mon, 2009-05-18 at 21:19 -0600, LuKreme wrote:
> On 18-May-2009, at 19:02, Michael Monnerie wrote:
> > I didn't mean that the final result be a FP, just this one ruleset.
> > Shouldn't the goal be to have no FPs and lots of corrects?
>
> In a word? No.
I don't think you understood what that DNSBL is about.
> Test are designed to be cumulative. Something that is seen 75% of the
> time in spam and 25% of the time in ham is still useful to give a
> positive (albeit small) score to. Something that is seen 99.7% of the
> time in spam and 0.3% of the time in ham is a lot more useful, so it
> gets a higher score. This doesn't mean we discard the first rule.
>
> The point of SA is not to have rules that ONLY hit spam or ONLY hit
> ham because the world does not work that way. If something is a
> potential spam indicator it can get a low score (0.1) or it can be
> used in combination with other rules to generate a rather high score.
Err, or maybe you just don't understand that we're not talking about REs
and rule-sets, but a blacklist.
--
char *t="\10pse\0r\0dtu...@ghno\x4e\xc8\x79\xf4\xab\x51\x8a\x10\xf4\xf4\xc4";
main(){ char h,m=h=*t++,*x=t+2*h,c,i,l=*x,s=0; for (i=0;i<l;i++){ i%8? c<<=1:
(c=*++x); c&128 && (s+=h); if (!(h>>=1)||!t[s+h]){ putchar(t[s]);h=m;s=0; }}}
