> It was good for a laugh, really. :) Until a strange feeling crept over
> me, realizing the words...
>
> Confirmed. That paragraph *severely* affected Bayes for me. No Bayes
> training with that mail up to that point, results reproducible.
>
> BAYES_99 probability 1.0000 WITHOUT that text
> BAYES_50 probability 0.5905 with that poisonous snippet
Got a second, almost identical one. After learning both, Bayes is back
to speed. The first, original spam including that snippet now also
scores BAYES_99 with a probability of 1.0000.
Oh, and the text actually is incorrect. Neither has been sent from the
USA, though it actually are dial-ups. ;)
--
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; }}}
