On 6/20/08 10:35 AM, Eliot Lear allegedly wrote:
William Herrin wrote:
Placing your mail server behind a NAT box and changing the NAT IP
during a move does absolutely nothing to update the public and private
RBLs and whitelists.
This case probably deserves more elaboration. Today, just about every
email reputation system is based on the IP address of the sender. There
are many such systems, including RBLs, PBLs, etc. Legitimate senders
that originate large volumes of mail do not like this precisely because
it takes time to develop a positive reputation. In some cases these
guys end up either inheriting or even sharing IP addresses through
common MTAs where their reputations are tied together.
Within the IETF we've developed DKIM that I hope will spawn reputation
systems based on domain and not IP addresses. However, that is very
much a speculative notion, and certainly far in the future. One reason
it's speculative is that it may well be that people want to use both IP
address AND domain to make their processing decisions. They can make
the answer on IP address far earlier in an SMTP transaction than they
can on domain, and that has its advantages.
The key point for this group is simply this: the IP address *is* an
identifier, an index key, and many other things that we'll never fully
know. It is possible to change, but only gradually over time. If it
doesn't have to, all the better.
Whats your conclusion? That we need to assume the IP address will be
used as an identifier by higher layers for the foreseeable future, and
figure out how to live with it?
--
to unsubscribe send a message to [EMAIL PROTECTED] with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg
