W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > September to December 1997

Re: draft-lundblade-1pass-mult-alt-00.txt vs. alternates

From: Graham Klyne <GK@acm.org>
Date: Wed, 12 Nov 1997 23:01:03 +0000
Message-Id: <3.0.32.19971112151236.007ad8f0@pop.dial.pipex.com>
To: Larry Masinter <masinter@parc.xerox.com>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
At 19:35 11/11/97 PST, Larry Masinter wrote:
>Is this a different model for 'alternates'?

I think so.

Since my first response to this draft, I have thought a little more...

The main point I would make is that the "types:"  proposal (per Lundblade)
lists *only* the MIME content-types which are present.

OTOH, the "alternates:" proposal also allows other negotiable features to
be incorporated into the list of options through the extension mechanism.
Specifically, it reserves "features" for the kind of negotiable feature
described in Koen's TCN draft and its progeny.

Which leads me to the question:

Can the 'alternates' proposal be used as a MIME 'multipart/alternate'
header in the fashion suggested by Lundblade?

I think the answer hinges mainly on the use of "absolute URIs" in the
'alternates' draft, which seems (at best) rather awkward when trying to
refer to body parts within the same MIME entity.  But I am not well
qualified in this matter.


(As an aside, for the 'alternates' header it might be worth making an
explicit reference to the Freed/Moore RFC 2231 (?) describing multiline
MIME headers.  It seems to be a good match.)

GK.
---

------------
Graham Klyne
Received on Wednesday, 12 November 1997 14:03:56 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:33:02 EDT