Re:

["Jack De Winter" <jack@xxxxxxxxxxxxxx>]

>In order to reduce code that means you want the envelope (MAIL
>FROM, RCPT TO) to be identical to ESMTP's envelope format.  And then your
>submission protocol ends up looking almost identical to SMTP.
>
>So I think the submission protocol should probably simply be a profile of
>ESMTP on a different port with different header-munging rules.  The submit
>draft does a good job of identifying the header-munging rules, but I'm
>concerned that the fact it permits use of those rules on the SMTP port via
>private agreement will block concensus.

I agree with this one, from a practicality point of view.  Submission
has certain advantages, and those can be honed for a particular
system.

What I am wondering is:
1) can we use the SUBMIT ESMTP extension that was dropped a while ago?
   this could be an alternative way to submission off of a submission port.
2) how practical are separate submission ports? can most servers handle
   it?  can mail clients today be set up to submit to a different port?
3) given that some mail servers do and some mail server's don't, could
   we define the VERIFYRCPT extension so that for ESMTP servers, this
   would state whether or not the default was to verify the addresses,
   and allow for the submitor to change that for their session?

The main concerns about all of these are stuff like: how to limit usage
to specific people (IP filtering is one way, authentication another),
default settings, and overhead on the system.

hoping to hear a lot of good stuff on this,
Jack

p.s. I think Randy's submission is a good start at #2.  It is concise
     and addresses the need now... as long as you are able to configure
     you mail system to run on two ports at the same time.
-------------------------------------------------
Jack De Winter - Wildbear Consulting, Inc.
(519) 576-3873		http://www.wildbear.on.ca/

Author of SLMail(95/NT) (http://www.seattlelab.com/) and other great products.