Message-Id: <3.0.1.32.19980330175006.007e22d0@lacroix.wildbear.on.ca>
Date: Mon, 30 Mar 1998 17:50:06 -0500
To: "Eric S. Raymond" <esr@snark.thyrsus.com>,
        Michael A Quinlan <mikeq@primenet.com>
From: "Jack De Winter" <jack@wildbear.on.ca>
Subject: Re: X- fields
Cc: DRUMS WG <drums@cs.utk.edu>
In-Reply-To: <19980317092724.04769@snark.thyrsus.com>
References: <004401bd51a5$8dfc0740$0122a5ce@mikeq>
 <004401bd51a5$8dfc0740$0122a5ce@mikeq>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

At 09:27 AM 3/17/98 -0500, Eric S. Raymond wrote:
>Michael A Quinlan <mikeq@primenet.com>:
>> My questions really are, 1) What is the current practice, and 2) should
this
>> be documented in the draft?
>
>The current practice is to use X- headers for private and some experimental 
>headers.  Dave Crocker asserts for good reason that the latter should be
>deprecated.  Robert Elz wants the former deprecated too.
>
>I would like X- headers for private use to be documented to avoid namespace
>clutter.  Current practice with them amounts to a declaration that the
>originator does not wish to step into the public namespace.

So for entering into this at this late date, but would registering
a vendor space help to resolve the clutter?  for example, if UW in its
imap server wanted to use X-IMAP, it could use X-V-UW-IMAP or something
like that... .perhaps V-UW-IMAP.  This would embed that it is a vendor
specific namespace, which vendor, and the extended header.

regards,
Jack
-------------------------------------------------
Jack De Winter - Wildbear Consulting, Inc.
(519) 884-4498		http://www.wildbear.on.ca/jacks/