Message-Id: <3.0.32.19971005110825.009d7a10@POP.Dial.Pipex.Com>
Date: Sun, 05 Oct 1997 17:09:36 +0100
To: Keith Moore <moore@cs.utk.edu>
From: Graham Klyne <GK@ACM.ORG>
Subject: Re: wishful thinking vs. reality 
Cc: Dave Crocker <dcrocker@brandenburg.com>, drums@cs.utk.edu
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

At 01:35 AM 10/3/97 -0400, Keith Moore wrote:

>Because the reply-to protocol is ultimately used by humans, not machines.
>We need for machines to behave by well-understood rules, but
>we expect humans to make choices and be responsible for those choices.
>
>(and as for replys from robots, there's a pretty good case to be made
>that they should always go to the return-path address...but if people
>disagree, let's not get into a discussion about that one right now...
>it's not on the table for drums)

This suggests to me some lack of clarity about the scope of a specification
whose title is "Message Format Standard", and whose introduction starts
with: 'This standard specifies a syntax for text messages that are sent
between 
computer users, within the framework of "electronic mail" messages...'

The introduction also says:
[...] 'This standard applies only to the format and some of the semantics
of message contents.' [...]
[...]
'Note: This standard is NOT intended to dictate' [...] 'any of the
characteristics of user interface programs that create or read messages.'
[...]

The current debate goes well beyond mere *syntax*, and to the extent that
the standard is about just syntax I think debate about SHOULD and MUST is
somewhat arbitrary.

To the extent that the syntax description should also indicate how it is
intended that that syntax be used, I question whether any appeal to RFC2119
for this material is appropriate.

I think this debate becomes meaningful only if the protocol document is
specifying the behaviour of User Agents (and Users?).  In which case, I
suggest the title and introduction of the document should be updated:
maybe if the scope of the document is clearly stated, it is easier to
decide what forms of mandated behaviour can be regarded as "in scope".

Finally, I think the point about any *user* being ultimately responsible
for any message (including any reply) they send is well made.

GK.
---

------------
Graham Klyne
GK@ACM.ORG