Re: [wsrp] [wsrp-wsia] [change request #245] UserScope and caching

[Alejandro Abdelnur <alejandro.abdelnur@sun.com>]

Title:

I've overlooked the consumerUserScopes element in the RegistrationData.
I withdraw the change request.

Alejandro

Rich Thompson wrote:

The more general statement is that this relaxation is only appropriate if we are confident that the "perUser" scope is a proper subset of the semantics of all interesting scopes that are likely to be defined. I doubt this is true and therefore favor the statement as it currently is in the spec.

Another view on this is that since the Consumer explicitly tells the Producer what userScopes it supports, the Portlet/Producer specifying a different userScope is explicitly stopping the caching of the markup and might be notifying the Consumer of an interest in a particular userScope (presuming reasonable logging is happening).

Rich Thompson

Andre Kramer <andre.kramer@eu.citrix.com> 03/24/2003 07:10 AM

To:        wsrp@lists.oasis-open.org         cc:                 Subject:        RE: [wsrp] [wsrp-wsia] [change request #245] UserScope and cachin        g

Say I wanted to have a "perApplicationUser" scope that specified how markup is to be handled over performBlockingInteraction boundaries. With this proposal, I could only use this if consumers supplied "perApplicationUser" as one of the consumerUserScopes in RegistrationData on (optional in-band) registration. [And I will have to advertise this scope as available e.g. using a read-only registration property!]
 
regards,
Andre
-----Original Message-----
From: Rich Thompson [mailto:richt2@us.ibm.com]
Sent: 21 March 2003 20:21
To: wsrp@lists.oasis-open.org
Subject: [wsrp] [wsrp-wsia] [change request #245] UserScope and caching


Spec 0.92/Section 6.1.5/Page 30/Lines 43-47

Old text: "If the Consumer does not know how to process the specified userScope, it MUST NOT cache the markup."

New text: If the Consumer does not know how to process the specified userScope, it MAY NOT cache the markup or it SHOULD handled as *perUser* userScope."

Reasoning: I think it is to strong to require not to do caching at all when the perUser scope would be a safe fallback as it is private to the user.