RE: workspace property

Just forbidding the MOVE is certainly reasonable behavior from a server
(that's what our server will do :-).  There is precedent for having a
property that is modified by a DELETE (e.g. the "DAV:parent-set" of a
resource that supports multiple bindings), but one certainly would
like to minimize the number of properties that act this way.

Before banning a cross-workspace move in the protocol though, I'd probably
prefer to wait a bit to make sure that this is in fact something that
does not turn out to be generally useful/supported.

Cheers,
Geoff



-----Original Message-----
From: Julian Reschke [mailto:julian.reschke@greenbytes.de]
Sent: Tuesday, October 01, 2002 3:05 PM
To: Clemm, Geoff; ietf-dav-versioning@w3.org
Subject: RE: workspace property


Not convinced yet :-)

The idea of a resource property which changes it's value upon MOVE isn't
really appealing -- the properties of a resource should be independant of
the access path.

Up to now, I was expecting a MOVE to have the same properties as a BIND
followed by a DELETE on the original URI -- this would be lost with your
interpretation.

Question: what is the use case for MOVEs between workspaces? Wouldn't it be
much simpler to just forbid that?

--
<green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760 
-----Original Message-----
From: ietf-dav-versioning-request@w3.org
[mailto:ietf-dav-versioning-request@w3.org]On Behalf Of Clemm, Geoff
Sent: Tuesday, October 01, 2002 8:30 PM
To: ietf-dav-versioning@w3.org
Subject: RE: workspace property


No, the DAV:workspace is not affected by the request-URL that you 
use to identify the URL (that would be bad for a variety of reasons). 
The only way you can have two different URLs for the same resource is 
if you have two bindings to either the resource or to a parent of the 
resource.  In this case, some resource has multiple parents, and which 
parent is picked for inheritance of the DAV:workspace property is up to 
the server (but it has to pick one, and return it consistently). 
So the answer was: (d) None of the above (:-). 
Cheers, 
Geoff 
-----Original Message----- 
From: Julian Reschke [mailto:julian.reschke@greenbytes.de] 
Sent: Tuesday, October 01, 2002 1:49 PM 
To: ietf-dav-versioning@w3.org 
Subject: DAV:workspace property 



Hi, 
a very basic question... 
The workspace feature introduces a live (protected) resource property which 
seems to depend on the URI of the resource, not the resource itself: 
" If the request-URL did not identify a workspace, the DAV:workspace of the 
destination MUST have been updated to have the same value as the 
DAV:workspace of the parent collection of the destination." 
To me this seems to indicate that it's not really a property of the resource

itself, but that it only depends on how you access the resurce. 
How does this property affect bindings to a resource in a workspace? 
a) there may be no additional bindings 
b) there may be additional bindings, but they must all reside in the same 
workspace's namespace 
c) there may be additional bindings, and the reported DAV:workspace property

just depends on the URI used in the request 
(I currently lean towards c) 


Julian 
-- 
<green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760