- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Tue, 30 Mar 2004 09:10:35 +0200
- To: Jason Crawford <ccjason@us.ibm.com>
- Cc: nnw3c-dist-auth___at___w3.org@smallcue.com
Jason Crawford wrote: > > The "hidden" flag you suggest would also be interesting as a > > binding property, as was suggested in this Internet-Draft > > (also implemented in Microsoft servers): > > http://www.ics.uci.edu/~ejw/authoring/props/draft-hopmann- > > collection-props-00.txt > > Although I'm not sure if hidden is a good example, I don't have a > problem with some hypothetical property returning a different value > depending on which URL you use to reference it. In fact I wouldn't be > surprised if we eventually intentionally define a property that does > vary by URL. But we should be clear right now that all properties that > are resource based (which is basically everything at the time of the > writing) should not vary by URL and that future properties should not > vary by URL without a documented reason. Well, I think that properties SHOULD NOT vary on request URI (nor should the content), as this is clearly against the goals of the BIND spec. RFC2518 says that PROPFIND returns the resource's properties, and BIND speaks about having multiple URIs for the same resource. I think this is clear enough... If you need responses that vary with the request URI, you're IMHO not talking to the *same* resource anymore. Regards, Julian -- <green/>bytes GmbH -- http://www.greenbytes.de -- tel:+492512807760
Received on Tuesday, 30 March 2004 02:36:47 UTC