RE: Versioning implications for Referencing

Personally I prefer the use of a method here.  Setting a property to change
the behavior of the system is a bit strange.  We did start this way, but it
means that you must have namespace properties as well as resource
properties.  What do I mean by that.  Well, I argue that the mechanism must
be consistent even if you have a direct/indirect reference to another
resource.  Otherwise the client needs to do special processing.  That means
the PROPPATCH must be able to change the namespace URI's properties
independent of the target URI's properties.  It looks like Judith's team is
willing to do this for direct references.  However, what about indirect?  We
could say you can't do it.  OK, but you are still setting a property that
will effect other GET requests.  It feels a little odd.  Of course, this is
what we are doing with the "default" workspace configuration.  So we either
need to change that to PIN or switch both to properties.  It just seems
cleaner to have a method and let the server do what is necessary to make the
right thing happen.

Chris

		-----Original Message-----
		From:	Geoffrey M. Clemm [mailto:gclemm@tantalum.atria.com]
		Sent:	Tuesday, November 03, 1998 7:09 PM
		To:	w3c-dist-auth@w3.org
		Subject:	Re: Versioning implications for Referencing


		I agree with the choice of a special request header.

		I believe it is likely that the "PIN" method will be
replaced with a
		property update, so the request header that would make
GETPROP and
		PUTPROP refer to the reference itself rather than the
reference target
		is a preferable (in addition to being more general) solution
than any
		special handling of a "PIN" method.

		Cheers,
		Geoff

		   From: "Slein, Judith A" <JSlein@crt.xerox.com>
		   Date: Tue, 3 Nov 1998 09:28:11 -0500 

		   My preference would be for the latter:  by default,
		   any method on a direct reference would be passed through,

		   but some header on the request would make the method
affect 
		   the reference itself.  I think we would still say that
DELETE, 
		   MOVE, and COPY always affect the reference, never its
target.

Received on Monday, 30 November 1998 14:30:44 UTC