Mime-Version: 1.0
Message-Id: <v04204d3eb371cd7747f6@resnick2.qualcomm.com>
In-Reply-To: <8210.927596454@munnari.OZ.AU>
References: <8210.927596454@munnari.OZ.AU>
Date: Wed, 26 May 1999 11:18:01 -0500
To: Robert Elz <kre@munnari.oz.au>
From: Pete Resnick <presnick@qualcomm.com>
Subject: Re: Take 2: References/In-Reply-To replacement text
Cc: DRUMS <drums@cs.utk.edu>
Content-Type: text/plain; charset="us-ascii" ; format="flowed"

On 5/25/99 at 11:40 AM +1000, Robert Elz wrote:

>s/an/a/

Oy. Done.

>  | domain name (or a domain literal IP address) of the host on which the
>
>s/)/, provided the IP address concerned is known globally unique)/

It's just a "good method" here, so I think the extra verbiage is 
unnecessary at this point.

>add
>
>It is best to use a representation (which need not be comprehensible
>by humans) of the date expressed in UTC here, to avoid the slight
>potential for problems should a domain name or IP address move to a
>different timezone.

A little wordy, but I'll throw in something to that affect.

>  | Though other algorithms will work, it is
>  | RECOMMENDED that the right hand side contain some domain identifier
>  | (either of the host itself or otherwise) such that the generator of
>  | the message identifier can guarantee the uniqueness of the left hand
>  | side within the scope of that domain.
>
>Why is this just recommended?   I always thought that was the whole
>point of the message-id syntax, and that nothing else has a hope of
>working.   The domain name used must be one in which the generator of
>the message-id can guarantee that the local part (of the message-id) is
>unique (so that the whole thing is then unique).   That can be a host name,
>or a higher level domain name (or even a domain literal (IP address) though
>that is harder to do correctly).  I had never conceived of the domain name
>part of the message-id being something about which there was any debate
>or difference of opinion.  Is there?

Remember, RECOMMENDED is equivalent to SHOULD; that is, do it unless 
you've got some good overriding reason not to. So for example, the 
rhs could be a faked-up identifier if someone chose to do that: 
"message-id.munnari.oz.au" might be reasonable even if it is not a 
registered domain name and doesn't correspond to any particular 
machine. There are all sorts of interesting circumstances where 
getting a real domain name (or IP address) might be a real pain in 
the butt.

pr
-- 
Pete Resnick <mailto:presnick@qualcomm.com>
Eudora Engineering - QUALCOMM Incorporated
Ph: (217)337-6377 or (619)651-4478, Fax: (619)651-1102