The producer could offer a degraded
representation of the resource say by the convension that the wsrp-url ==
resourceID is formatted as a URL if http, say, is an alternative fallback. And
not rendering the resource is a tactic often encountered on the Web.
Regards,
Andre
From: Rich Thompson
[mailto:richt2@us.ibm.com]
Sent: 22 December 2004 16:13
To: wsrp@lists.oasis-open.org
Subject: RE: [wsrp] Issue #29: Is
metadata needed to indicate support forgetResource?
I think it is dangerous to suggest an http proxy
request for a resource that specified using getResource is likely (or expected)
to work. The references to state, etc for handling the call are totally
different. Even if the Producer provides a proxy facility for its portlet's
resources, how is the Consumer to know the correct url (either base or format
of a reference or querystring)?
Rich
Andre Kramer
<andre.kramer@eu.citrix.com>
12/22/2004 10:24 AM
|
To
|
"'Michael Freedman'"
<Michael.Freedman@oracle.com>, wsrp@lists.oasis-open.org
|
cc
|
|
Subject
|
RE: [wsrp] Issue #29: Is metadata needed to
indicate support forg etResource?
|
|
forgetResource (subject)? Forward or backwards
interop can be by degrading the HTML or falling back to HTTP but I would
encourage 2.0 (consumer) implementations to support getResource and 2.0
producers to also provide a 1.0 service.
Regards,
Andre
From: Michael Freedman
[mailto:Michael.Freedman@oracle.com]
Sent: 21 December 2004 17:55
To: wsrp@lists.oasis-open.org
Subject: Re: [wsrp] Issue #29: Is metadata needed to indicate
support forgetResource?
And vice-versa: so a producer can adapt to run in a 1.0 consumer.
Rich Thompson wrote:
Issue # 29
Spec section:
SubCommittee: Interfaces
Owner: Andre
Description: Isn't
metadata needed to tell a Consumer whether or not a Producer supports
getResource (or is it always supported if urls encode for it and Consumers
never invoke it unless a url has encoded wsrp-useOperation=true)?