- From: Roy T. Fielding <fielding@avron.ICS.UCI.EDU>
- Date: Tue, 29 Nov 1994 23:43:57 -0800
- To: Albert Lunde <Albert-Lunde@nwu.edu>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Albert Lunde writes:
> Is use of MIME or MIME-like multi-part types a current practice
> supported by either clients or servers? (My impression was that it was
> not used and it could confuse a lot of clients.)
MIME multipart types have been in the old spec for over a year. Although
the major public servers and clients have not implemented them, some
private implementations have.
> If not, I'd suggest removing it from this version of the
> specification. I note the disclaimer that this specification
> may not reflect current practice, but still, codifying current
> practice is a main aim of the HTTP/1.0 spec.
It is possible (and anticipated) that some of what is in the current draft
will be considered "not compatible with current practice" and be shoved
over to HTTP/1.1. Henrik and I decided that we needed to include everything
necessary to implement what was incompletely specified in the old draft.
For example, support for PUT requires a bunch of stuff which has not yet
been implemented within a public web server, and yet which is known to be
necessary to support group annotation. Since it is much better for
committees to decide "what needs to be removed" than it is for them to
decide "what needs to be included", I think this is the best route to take.
I anticipate this will be a large part of the discussion in San Jose.
......Roy Fielding ICS Grad Student, University of California, Irvine USA
<fielding@ics.uci.edu>
<URL:http://www.ics.uci.edu/dir/grad/Software/fielding>
Received on Tuesday, 29 November 1994 23:53:04 UTC