Re: [wsrp-wsia] Re: [wsrp-interfaces] RE: [wsia] Sessions

[Rich Thompson <richt2@us.ibm.com>]

I would agree that this is a good writeup from a portlet developer
perspective. Basically arriving at 2 very reasonable conclusions; 1) Most
portlets want to view their "session-data" as private and in their own
session. 2) When a portlet developer will be depending on sharing their
"session-data", a mechanism outside the scope of WSRP/WSIA wants to be
employed (basically the producer service enabling the appropriate sharing).
In either case, there is a need to carry the equivalent of the cookie since
the underlying infrastructure doesn't automatically carry it for us. Also,
this analysis ignores the case we have been talking about where the
Consumer wants to cause data to be shared between portlets that weren't
developed with a dependency on this sharing. There can be significant value
in doing this and it is only when the Consumer aggregates the portlets that
it is reasonable to define the groups that will be involved in such data
sharing.



                                                                                                                    
                      Alejandro                                                                                     
                      Abdelnur                 To:                                                                  
                      <alejandro.abdeln        cc:       wsrp-interfaces@lists.oasis-open.org,                      
                      ur@sun.com>               wsrp-wsia@lists.oasis-open.org                                      
                                               Subject:  [wsrp-wsia] Re: [wsrp-interfaces] RE: [wsia] Sessions      
                      06/17/2002 12:29                                                                              
                      PM                                                                                            
                                                                                                                    
                                                                                                                    



Eilon,

Good write up. Some comments follow.

* Sessions in Browser.

The assumption "multiple browser windows equate to one set of cookies" is
true when all the browser windows are part of the same process. While this
is a common case in Windows it does not necessary holds true, MS Explorer
can be configured to use separate processes for each browser window, in
that case each browser window has its own set of cookies.

* Basic Scenario.

I agree with your conclusion. I would just state that the consumer has to
provide to the producer enough information so the producer knows that 2
requests are meant for the same client, this would allow the producer to
use the same session for both request if it needs to do so.

* Sharing between Groups Portlets and Sharing between Group and User
Portlets

I see these 2 cases outside of the scope of the WSRP, at least for the
first version.

Regards.

Alejandro

Eilon Reshef wrote:
      Carsten and team,

      Attached are some additional thoughts on sessions - in particular how
      people may want to use sessions in a different way that I am not sure
      works very well with our current approach.

      Eilon
            -----Original Message-----
            From: Carsten Leue [mailto:cleue@de.ibm.com]
            Sent: Friday, June 14, 2002 11:16 AM
            To: wsia; WSRP; wsrp-interfaces@lists.oasis-open.org;
            wsrp-wsia@lists.oasis-open.org
            Subject: [wsia] Sessions




            As promised (though a bit late) here is a sketch of my current
            understanding of the notion of sessions.





            (See attached file: Sessions.zip)


            Best regards
            Carsten Leue


            -------
            Dr. Carsten Leue
            Dept.8288, IBM Laboratory Böblingen , Germany
            Tel.: +49-7031-16-4603, Fax: +49-7031-16-4401