RE: Get on a reference

Judy, John, Jim:

How about this:

(1) GET on a reference returns 302 as per John's proposal.
(2) PUT works on a reference, unless what is being put
is not a reference, in which case it fails.

Then, browsers would work, clients could control whether
they are dealing with the reference or the target both for
read and writing.

Alan Babich
"Cogito ergo zoom." (I'm a Corvette ZR1 owner -- 405 hp, 180 mph.)

> -----Original Message-----
> From:	Judith Slein [SMTP:slein@wrc.xerox.com]
> Sent:	June 30, 1998 6:46 AM
> To:	Jim Davis
> Cc:	w3c-dist-auth@w3.org
> Subject:	Re: GET on a reference
> 
> At 03:19 PM 6/29/98 PDT, Jim Davis wrote:
> >At 08:22 AM 6/29/98 PDT, Judith Slein wrote:
> >>I agree that this is the direction the discussion in Redmond seemed
> to be
> >>pointing.
> >
> >>There is a requirement (3.1.4) that operations on a referential
> member not
> >>affect the resource it references.... If requests were automatically
> >redirected to its target resource,
> >>this would not be possible.
> >
> >There is nothing automatic about the 302 response.  A client can
> decide
> >whether to chase the link or not.  And so what if it is automatic?
> By
> >hypothesis, the client does not understand DAV, so there's no choice
> it can
> >make.  Is there any operation that John's proposal makes impossible?
> I
> >think not.
> 
> I was not assuming that the client doesn't understand DAV.  A client
> that
> does understand DAV may have occasion to do a GET or HEAD on a
> referential
> resource in order to see the headers for the referential resource.  Do
> those headers come back with a 302 response? (I was assuming not.)  If
> not,
> then how does the client ever get access to them?
> 
> >
> >>Similarly for PUT, the client might be wanting to replace the
> content of
> >>the target resource, but on the other hand it might be wanting to
> replace
> >>the reference with an ordinary resource that has content. 
> >
> >I don't think John proposed that a PUT do redirection.  Don't we
> already
> >say that PUT on a referential resource is an error, just as PUT on a
> >collection is an error?
> 
> At the moment the spec says that the referential resource will be
> replaced
> by whatever was in the PUT.  (That is, it will not be redirected, and
> it
> will not fail.)
> 
> >
> >It's not even clear to me it would do anything.  (e.g, if you do a
> PUT with
> >Netscape and it gets a 302 response does it retry the PUT at the new
> URL?)
> > 
> >I see nothing to object to in John's proposal.
> >
> >
> >
> Name:			Judith A. Slein
> E-Mail:		slein@wrc.xerox.com
> Internal Phone:  	8*222-5169
> Fax:			(716) 422-2938
> MailStop:		105-50C
> Web Site:    http://www.nde.wrc.xerox.com/users/Slein/slein.htm

Received on Wednesday, 1 July 1998 16:33:21 UTC