On Fri, Oct 22, 1999 at 01:45:09PM -0500, Chris Garrigues wrote:
> Digitally signed messages containing the SPAM addresses and you have to
> submit your public key before you can send the SPAM addresses; abusers would
> get rapidly removed from the trusted list.
That is good. If you can identify who is submitting the bad
addresses then abuse can be controlled much easier.
> Also, an address doesn't get added until two (three? twelve?) different sites
> report it as a SPAMmer. Since real SPAMmers, by definition, hit everywhere,
> that shouldn't keep real SPAM from being blocked quickly, and would make it
> a bit harder (when combined with my above suggestion) to compromise the system.
Maybe is should be up to the users of the database to decide how
many times a site must be submitted.
You could make the database complex and have different reasons
for the site being in the database (open relay, in Vixie db, sent
mail to spambait address, etc.). The clients could determine
what constitues being a "spam site".
Mechanism is should be concentrated on rather than policy.
Policy is a site decision.
Neil
