Unless you want to discuss implementation details and work on this
yourself, the users list is the appropriate place.
On Fri, 2009-01-09 at 03:21 -0800, an anonymous Nabble user wrote:
> rules like:
>
> -Same subject already on n previous mails
> -Same subject already on n previous mails on x hours
> -Same IP send previous mail from distinct "From"
[...]
> already can be scored "similar" subjects, bodys, urls on body...
>
> I know is need a big and fast database of "source ip", "Froms", "Subjects",
> "body summary"... of all mails received on x days.
>
> If a lot of servers can share this database spam can avoid before receive
> first spam mail. But must be solved how to ensure "privacy" of this data.
>
> exist something like this? i search it but i not found
AFAIK, no plugin / service exists a you describe it above. And yes, that
likely will require a huge database and implies quite some privacy
issues.
Digest services are what comes closest, like Razor, Pyzor, iXhash, DCC.
However, they do not use verbatim headers, senders and IPs, but
calculate "fuzzy" checksums by removing some of the per-spam noise
injected specifically to evade strict, verbatim comparison of single
headers.
guenther
--
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; }}}
