Message-ID: <8BB128650D00D211BAEF00A0C9C74EB902EA07@PILGRIM>
From: "Larry Osterman (Exchange)" <larryo@exchange.microsoft.com>
To: "'John C Klensin'" <klensin@mci.net>
Cc: drums@cs.utk.edu
Subject: RE: RE: DATA/RSET/NOOP arguments
Date: Thu, 30 Jul 1998 11:24:13 -0700
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"

But you DID do ABNF for the 250 reply code (ehlo-ok-rsp)......  

My feeling is that if there are defined protocol elements in a response,
that needs to be in the spec.  So if the 220 response is REQUIRED to have a
domain before the greeting string, it should be in the spec....

If, on the other hand, the "220 <domain> Service ready" response is just
convention and is not required for interoperability, it clearly doesn't need
to be specified.

As far as I can tell, this only applies to 220, 221 and 421, the other
responses appear to be simply text.


Larry Osterman
Sent from larryo-laptop.dns.microsoft.com running NT5 and Outlook 98 and
Exchange Server 5.5.  Please notify the sender of any difficulties


-----Original Message-----
From: John C Klensin [mailto:klensin@mci.net]
Sent: Thursday, July 30, 1998 11:17 AM
To: Larry Osterman (Exchange)
Cc: drums@cs.utk.edu
Subject: Re: RE: DATA/RSET/NOOP arguments


On Thu, 30 Jul 1998 09:36:33 -0700 "Larry Osterman 
(Exchange)" <larryo@exchange.microsoft.com> wrote:

> Oh, and before you complain, yes, I am aware of existing practice that
scans
> for the string ESMTP in the 220 line, in MY honest opinion, that is one of
>...
> Actually, on rescanning the document, I don't see any ABNF that describes
> the 220 line, except for the following:
> 
> 	220 <domain> Service ready
> 
> Shouldn't there be grammar describing the greeting message as well?
> Something like:
> 
>      welcome      ::= "220"    domain [ SP greeting ] CRLF
>      greeting     ::= 1*<any character other than CR or LF>
>...

We haven't done ABNF for any of the reply codes, partially 
because it would have required some very circuitous 
language to permit the use of RFC 1893 coding without 
forcing a normative reference to 1893 and/or forcing us to 
debate whether that format should be required at this time.

Is there a strong feeling that can of worms should be 
opened?

    john