Ian Eiloart wrote: > > > --On 22 July 2006 20:23:32 +0800 W B Hacker <[EMAIL PROTECTED]> wrote: > >> >> No point seen in looking at a 'yellow' list, of servers that sometimes >> send spam and sometimes do not. By definition, that should be most of >> the world. >> > > Yes, but knowing that a host has been seen sending good email is useful > information - it might make you less likely to reject. > > Perhaps more useful for bayesian analysis would be a result that encoded > the proportion of seen mail that was ranked as spam. For example, > returning an IP address 127.0.0.xxx where xxx varied from 0 for no spam > to 255 for no ham. > >
Bayesian pays its dues on a per-account MUA. We had not found it worth the electricity on a multi-user, multi-domain server. OTOH, given that we typically do 80-85% of our rejection before even calling a stripped-down SpamAssassin, there wouldn't be much left for Bayes to analyze even if we still ran it. Bill -- ## List details at http://www.exim.org/mailman/listinfo/exim-users ## Exim details at http://www.exim.org/ ## Please use the Wiki with this list - http://www.exim.org/eximwiki/
