RE: [wsrp-interfaces] Tomorrow's agenda

["Andre Kramer" <andre.kramer@eu.citrix.com>]

I think a larger question is whether or not a WSRP Consumer and "AJAX"
portlets sharing a page can leverage the current protocol (where the new
getResource should help).

A consumer could also avoid a full page reload by use of the 1.0 cache
control validateTag. For a scripted page, if an action and all following
events result in no change in validate tags returned then the consumer
(and action-ing portlet) could effectively "skip" the getMarkup phase.

This "no need to refresh page" (the inherent good or bad of AJAX -
depending on your viewpoint :-) would have to be indicated to the
"dynamic" portlet that requested the action (or the event processing)
but that may well be a private contract between the portlet(s) and the
consumer, implemented by consumer markup re-writing (assume re-writing
is needed to handle other complexities of scripting).

If so, the JSR (or others) could standardize the consumer/portlet
contract and WSRP 2.0 committee draft could be used largely as is
(depending on your interpretation of caching and the requirement to
re-write URLs used by scripts). I won't make the call but I think an
investigation of such a use-as-is hypothesis is worth active
investigation?

Regards,
Andre

-----Original Message-----
From: Subbu Allamaraju [mailto:subbu@bea.com] 
Sent: 02 March 2006 01:52
To: wsrp-interfaces@lists.oasis-open.org
Subject: Re: [wsrp-interfaces] Tomorrow's agenda

 From my recollection and understanding, the key question still open is 
whether getResource is the recommended approach for this version. For 
example, one of your emails recently questions the suggestion for 
implementations to use getResource instead of extensions.

We need to send a clear signal on this topic, particularly in the wake 
of the latest feedback we got in the comment list.

Subbu

Michael Freedman wrote:
> Can you clarify what you want to discuss?  The minutes for the last TC

> concall indicates we decided to defer support beyond what can already
be 
> accomplished using getResource until 3.0.  Yes, this topic can be 
> reopened but wouldn't it be better to do at a TC call as that was
where 
> the discussion/decision was made?    -Mike-
> 
> Subbu Allamaraju wrote:
> 
>> Mike,
>>
>> Could you get the question of Ajax onto the agenda please? While 
>> discussing some of the topics below in the email thread, ajax was 
>> mentioned several times, and I suspect we will end up doing the same 
>> tomorrow.
>>
>> Subbu
>>
>> Michael Freedman wrote:
>>
>>> I have attached a document that lists 3 open discussion items from 
>>> the wsrp list + the security topic.  I suggest we spend a little
time 
>>> on the open issues but also make sure we have enough time to begin 
>>> the security discussion.  The topics have links to the mail threads 
>>> so you can refamiliarize yourself.
>>>    -Mike-
>>>
>>>
------------------------------------------------------------------------
>>>
>>> navigationalParameterDescriptions: 
>>>
<http://www.oasis-open.org/apps/org/workgroup/wsrp/email/archives/200602
/msg00027.html> 
>>>
>>> Do we need to clarify language concerning consumers recognizing 
>>> like-named navigational parameters as related/auto wired?
>>>
>>> Clarification: Resource URLs and markup params, etc. 
>>>
<http://www.oasis-open.org/apps/org/workgroup/wsrp/email/archives/200602
/msg00030.html> 
>>>
>>>
>>>    1. Stateless consumer
>>>    2. Meaning of title?
>>>    3. Change window state/mode?
>>>    4. Concurrency problems with sessionID?
>>>    5. During *getMarkup, handleEvents *and
*performBlockingInteraction*
>>>       invocations the Consumer indicates to the Portlet its current
mode
>>>       via the MarkupParams data structure"  Add getResource???
>>>    6. Clarity of structure vs reuse of structure -- Should we have
>>>       specific types to clarify that a few fields are no sensible?
>>>    7. Caching when using getResource of portlet markup ???? Not yet
>>>       raised ---
>>>
>>> unclear behavior for dynamic nav params and transient properties 
>>>
<http://www.oasis-open.org/apps/org/workgroup/wsrp/email/archives/200602
/msg00076.html> 
>>>
>>> Why don't we allow nav params/transient props to be introduced 
>>> dynamically like events?  Should we?
>>>
>>> Security questions: 
>>>
<http://www.oasis-open.org/apps/org/workgroup/wsrp-interfaces/email/arch
ives/200602/msg00003.html> 
>>>
>>> Can we recommend that all producers support at a minimum UserName 
>>> with no consumer authentication?  Can we encourage all consumer to 
>>> send UserName with no consumer auth unless otherwise instructed to
do 
>>> otherwise?  Will a producer work if it doesn't handle the WS stuff 
>>> but receives it?  I.e. will it just be ignored?
>>
>>
>>