From: Matt Curtin <cmcurtin@interhack.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <14164.50727.851481.146322@strangepork.interhack.net>
Date: Wed, 2 Jun 1999 01:50:31 -0400 (EDT)
To: Keith Moore <moore@cs.utk.edu>
Cc: "Larry Osterman (Exchange)" <larryo@exchange.microsoft.com>,
        "'Robert Elz'" <kre@munnari.oz.au>,
        Pete Resnick <presnick@qualcomm.com>, DRUMS <drums@cs.utk.edu>
Subject: Re: Message-ID (was: RE: Take 2: References/In-Reply-To replaceme nt text) 
In-Reply-To: <199906020424.AAA14744@astro.cs.utk.edu>
References: <14164.41068.812072.418764@strangepork.interhack.net>
	<199906020424.AAA14744@astro.cs.utk.edu>

>>>>> On Wed, 02 Jun 1999 00:24:57 -0400, Keith Moore <moore@cs.utk.edu> said:

>> like: <1999060122581400cmcurtin@interhack.net>

Keith> what operational purpose does this serve?

It's just an example of a "safe" generation scheme that would allow:
 o division of the namespace by Internet domain name.
 o calculation of the correct Internet domain name without having to
   use something that could potentially be an internal host name
   (which might be the case if we use a gethostbyaddr(), hostname,
   etc.)

On USEFOR, some complained that it's a very bad for Message-IDs to use
blindly the host-part of the sender's address, as extremely large
domains with a *lot* of clients out there (aol.com, hotmail.com, etc)
generating Message-IDs could collide with others in the same zone.

Sticking the user name at the end of the LHS would effectively
eliminate that complaint--though it wasn't clear that there was
consensus that it was something that had to be handled.

It does have some interesting and useful side-effects, like making
scoring easier...

Anyway, it was just an example, possibly more complicated than I
should have offered.

>> Should all of this Message-ID business be part of DRUMS itself, or
>> would it be possible to have a seperate document that details
>> Message-ID generation and usage, published with DRUMS?

Keith> probably the latter.

Alright.  I'll need to update my proposal in order to articulate
requirements for the Message-ID on the Internet more formally (i.e.,
RHS MUST be FQDN, as opposed to a casual "use of an unqualified
hostname...would be foolish").  I'll post it here and mail it to the
drafts editor when done (tomorrow?)...

-- 
Matt Curtin cmcurtin@interhack.net http://www.interhack.net/people/cmcurtin/