W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > January to April 1996

Re: remove PATCH, COPY, MOVE, DELETE, etc.? Upgrade?

From: Roy T. Fielding <fielding@avron.ICS.UCI.EDU>
Date: Fri, 23 Feb 1996 18:22:41 -0800
To: Larry Masinter <masinter@parc.xerox.com>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Message-Id: <9602231822.aa15115@paris.ics.uci.edu>
> I don't think that 'logical completeness' is a useful criteria for us
> to use in the decision of "does this go in HTTP/1.1 or put off for an
> extension or a later version".

I'll disagree with you there.  I think logical completeness is an
important part of any proposed standard, and I tend to oppose any
change until the change is complete.

> My reason for suggesting that DELETE, TRACE and WRAPPED not be in 1.1
> is my perception that
> 1)  there are likely to be disagreements about their form  (pure
> conjecture)

I claim that they are innocent until proven flammable.

> 2)  we've not discussed them (hypermail pointer to
> discussion/consensus about these in our mail archive?)

DELETE -> Why discuss it?  There is no controversy about it.

TRACE  -> http://www.ics.uci.edu/pub/ietf/http/hypermail/1995q4/0404.html
   with complete resolution of Max-Forwards proposal in

         http://www.ics.uci.edu/pub/ietf/http/hypermail/1995q4/0405.html

   That was one of the most peaceful debates we've had on any feature.  :)

WRAPPED -> Part of Dave Kristol's original extension proposal, with additions
   from myself and Rohit (for PEP).  I am not sure whether this qualifies
   or not.

> 3) there are no current implementations (pointers to existing
> implementations and their docuemntation?)

DELETE is implemented in libwww-perl.  Of course, since I designed it
to be extensible, I can implement any method in libwww-perl in less than
10 minutes.  I guess what you are looking for is a server implementation,
and I don't think that will happen until TRACE is in the proposed
standard.  However, if necessary, I'll spend some time implementing
it for Apache.

> 4) the form that these operations exist in within the HTTP/1.1
> specification has not had adequate review (list of HTTP-WG members who
> claim to have reviewed these sections carefully?)

That is not a valid question.  Had you formed a subgroup to do that,
there obviously would have been an "adequate review".  As it stands,
I have had no objections from anyone, inside or outside the WG,
regarding those methods (aside from the brief discussion of TRACE,
which was resolved to everyone's satisfaction).

I can live without WRAPPED, but removing TRACE and DELETE will create
more controversy than you seek to avoid by removing them.

......Roy
Received on Friday, 23 February 1996 18:29:55 EST

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:31:46 EDT