----- Original Message -----
From: "J C Lawrence" <claw(_at_)kanga(_dot_)nu>
To: "Hector Santos" <winserver(_dot_)support(_at_)winserver(_dot_)com>
If we assume a discrete HEAD/DATA distinction at the transport protocol
level, with a HEAD definition that supports a cheaply determined binary
quality: auditable/unauditable (for its audits we're really talking
about rather than traceability) then you both get what you want: Hector
gets the technical ability to (cheaply) refuse all email that has an
unauditable HEAD and to then deal with his auditable mail as he wishes,
and Len has the technical ability to choose which types of
auditable/unauditable email he accepts. Of course such an approach
requires auditable binding of HEAD data to DATA contents, but that is
tractable.
Hector: Would you have a problem with such an approach, and if so, why?
HEAD/BODY idea.
In general, I think it will work with the assumption all the information is
there to "audit/trace/track ____ fill in"
From a software design standpoint, a see HEAD/BODY transaction as easy to do
You still need an EHLO like concept unless we also have a "server attribute"
concept at the DNS lookup level.
Getting all your header in one shot might bring up some bandwidth questions
to investigate. Off the top of my head I will estimate an average of 10-15
lines, max 80 characters or <~ 1K per HEAD transaction. It has all the
major benefits of performing all the different validation concepts we can
think. It also allows for the introduction of new headers that better fits
the HEAD idea.
I personally don't see a problem with this. I think it is a good idea. Short
and sweet and does the job. Of course, all the possible
successful/rejection responses need to be considered and it must be
enforceable as a system policy by default.
I'm signing off this list now. No time for this and I don't work well with
threats and censorship cops watching my every word or questions (Sorry, I
really didn't know what MAIL-NG was or how consensus was met!) Instead? I
get a threat of banishment!
Its been fun people.
--
Hector Santos, Santronics Software, Inc.
http://www.santronics.com