On 3-feb-04, at 10:21, Hector Santos wrote:
Will NG-MAIL make DNSRBL methods obsolete?
Of course not. Blacklisting will always be a necessary part of any
messaging system. But hopefully most of the blacklisting can be done on
email address or domain rather than on mail server. But if there are
rogue mail servers, people will want to blacklist them. What we need in
a new generation system that we lack today is the capability to inform
the offending system of what's going on an what they can do about it.
Today the choices are mostly limited to accept and reject. Obviously I
don't want to accept something that's likely to be spam, but that
doesn't mean I necessarily want to reject it either. For instance, if
it's an end user who is trying to deliver the message himself, I might
want to say "have your ISP deliver it". For a regular user this isn't a
huge inconvenience (at least not compared with having the message
rejected) but for a spammer it is as no ISP in their right mind will
allow their mail server to relay a spam run with millions of messages.
Or I might want to reroute the message to a server that has better spam
filtering. Or have them try again later, so I have time to see whether
they turn out to be spammers.