Date: Mon, 27 Jul 1998 21:22:16 -0400 (EDT)
From: John C Klensin <klensin@mci.net>
Subject: Re: smtpupd-07 comments (fwd)
To: Chris Newman <Chris.Newman@innosoft.com>
Cc: Detailed Revision/Update of Message Standards <drums@cs.utk.edu>
Message-id: <SIMEON.9807272116.I@tp7.Jck.com>
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=US-ASCII
Priority: NORMAL

Everyone,

Sorry, I've been at INET and IFWP meetings and largely off 
email for the last week.   For everyone's reference about 
what is going on here, I have identified below the point at 
which I stopped reading Dan's message when it first 
appeared.  That information is followed by an editor's 
note on procedures; it would help move things forward 
quickly if its general guidelines were understood and 
agreed to.  If such agreement is not possible, I'll need 
advice from Chris as to how to proceed.

>... 
> ---------- Forwarded message ----------
> Date: Fri, 05 Jun 1998 06:53:41 +0000
> From: "D. J. Bernstein" <djb@cr.yp.to>
> To: drums@cs.utk.edu
> Subject: smtpupd-07 comments
> 
> > The objective of the Simple Mail Transfer Protocol (SMTP) is to
> > transfer mail reliably and efficiently.
> 
> Insulting the reader's intelligence is not a good way to start 821bis.
> SMTP is amazingly inefficient, and thanks to some protocol design errors
> it has proven to be terribly unreliable in practice.
> 
> Was Postel trying to compare SMTP to FTP? I assure you that this is of
> zero interest to today's implementors.
>...

Here I stopped reading since I do not believe my job 
description as editor requires putting up with this tone. 
This sort of approach makes it terribly difficult for any 
embedded technical comments to get through.

Editor's comment: I agree with Chris that DRUMS has dragged 
out too long.  I would add that there is evidence that 
exhaustion has set in to the extent that we need to worry 
about whether any consensus achieved reflects a real WG 
consensus or merely the burned-out agreement of those who 
have been able to put up with the duration and 
sometimes-tone of the efforts.

There were several conversations early in the process of 
putting 821bis together about the risks of making 
inadvertent changes to the intent of the specfication, 
thereby invalidating conforming implemations.  While I've 
tried to reorganize the material somewhat to make it easier 
to follow, I've also tried to be very conservative about 
changing text that didn't seem harmful, i.e., an "if it 
ain't broke, don't fix it" principle has often been 
applied.  I believe that principle becomes particularly 
important at this stage in the process as hasty changes 
after some perhaps-significant proportion of the natural 
membership of the WG has tuned out could be very risky.

So, I would like to suggest the following:

(i) Simple editorial corrections --e.g., typographic errors 
or ill-constructed sentences-- should be sent to me, 
ideally with suggested fixes.

(ii) Suggested editorial improvements --e.g., "remove 
paragraph X since it doesn't accomplish anything" or 
"reword paragraph Y" -- should be sent to the list, 
preferably with suggested replacement text.  The latter 
should help reassure everyone that we have reasonable 
consensus that the change is desirable before I start 
changing things.  And I would encourage people to evaluate 
such changes before suggesting them against the risk of 
dragging DRUMS out another meeting cycle.  It should be 
understood that, at this late stage, I may ignore editorial 
comments of this sort that do not come with suggested text 
if it isn't obvious to me what to do with them.

(iii) Substantive changes should go to the list and must 
come with suggested replacement text unless the way to make 
the change is trivially obvious.  It is just too late in 
the process for us to be guessing what each other are 
talking about.

And, finally, to be specific about some of the implications 
of my "stop reading" policy, people sending comments should 
simply assume that I won't read any message that contains 
ad hominem comments or attacks on the intelligence or good 
faith of any present or former participant or group of 
participants in the process.  I am hypersensitive in this 
regard: even the implication of such an attack, such as 
phrases like "have you forgotten" or "it's absurd" will 
cause a reflex action involving the delete key.

    john