wsrp message
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]
Subject: Re: [wsrp] Few clarifications on stateChange flag and consumer-managedstate
- From: Rich Thompson <richt2@us.ibm.com>
- To: OASIS WSRP TC <wsrp@lists.oasis-open.org>
- Date: Thu, 16 Mar 2006 10:55:39 -0500
Sure, modifyRegistration (7.3) returns
a RegistrationState (5.1.24) which does not include a registrationHandle
since the Producer is not allowed to change the handle when processing
this operation. This was also true in v1, though the section numbers were
different.
I understand that your only concern
is the discussion of 'readWrite'. My point was that 'cloneBeforeWrite'
is the only place where the semantics allow a clone to be returned. We
could state the negative for both of the other defined values for the portletStateChange
flag, but applying such an approach across the spec would become onerous.
On the other hand, if the consensus is that additional statements would
provide clarity, then we should add them.
Rich
Subbu Allamaraju <subbu@bea.com>
03/16/06 10:42 AM
|
To
| Rich Thompson/Watson/IBM@IBMUS
|
cc
| OASIS WSRP TC <wsrp@lists.oasis-open.org>
|
Subject
| Re: [wsrp] Few clarifications on stateChange
flag and consumer-managed state |
|
Rich Thompson wrote:
>
> The 2nd is not needed as the Producer CAN NOT (stronger than MUST
NOT!)
> return a registrationHandle from modifyRegistration.
Could you me to the section that has this statement? The producer has to
return the current registrationHandle since it is required in the
schema. My expectation is that the producer returns an updated
registrationState along with the current registrationHandle.
>
> The only setting of portletStateChange that allows for cloning is
> 'cloneBeforeWrite' (section 6.4.3). The PortletContext.portletState
> description explicitly connects the value supplied to the value in
the
> required portletHandle field. Personally this structurally states
the
> equivalent of the suggested conformance statement. However, if others
> think an additional conformance statement does make sense, I would
> suggest adding it to 6.4.3 (where the semantics of the
> portletStateChange flag are discussed).
Just want to make sure we are talking about the same thing. I agree that
the spec is clear on cloneBoforeWrite. The proposal is only for readWrite.
Subbu
> *Subbu Allamaraju <subbu@bea.com>*
>
> 03/16/06 09:38 AM
>
>
> To
> Rich
Thompson/Watson/IBM@IBMUS
> cc
> OASIS
WSRP TC <wsrp@lists.oasis-open.org>
> Subject
> Re:
[wsrp] Few clarifications on stateChange flag and consumer-managed
> state
>
>
>
>
>
>
>
>
> Rich,
>
> I agree with that interpretation. But the spec does not specify that
the
> producer MUST NOT return a new portletHandle when readRead is
supplied
> by the consumer.
>
> In the case of registration, the producer MUST NOT return new
> registrationHandle in response to a modifyRegistration request.
>
> I think both these are required since consumers may be using the handles
> as keys in databases.
>
> Subbu
>
> Rich Thompson wrote:
> >
> > PortletHandle is a required field in PortletContext and
the semantics
> > are that any portletState included in the PortletContext
is relative to
> > the portletHandle (Consumer MUST return it on future use
of the
> > portletHandle).
> >
> > I think this eliminates the need for the conformance statement
as it is
> > already encapsulated in the structure of the message.
> >
> > Rich
> >
> >
> > *Subbu Allamaraju <subbu@bea.com>*
> >
> > 03/16/06 08:34 AM
> >
> >
> > To
> >
OASIS WSRP TC <wsrp@lists.oasis-open.org>
> > cc
> >
> > Subject
> >
[wsrp] Few clarifications on stateChange flag and
> consumer-managed state
> >
> >
> >
> >
> >
> >
> >
> >
> > Mike Caffyn and I had some discussions on the correct interpretation
of
> > the stateChange flag when the consumer is managing the
state (via
> > portletState).
> >
> > Here are the possibilities:
> >
> > Consumer: Sends a readOnly flag and portletContext
> > Producer: Can either throw a fault if changes are required,
or process
> > normally but must not return a portletContext.
> >
> > Consumer: Sends a cloneBeforeWrite flag and portletContext
> > Producer: Returns a portletContext with a new portletHandle
and possibly
> > new portletState
> >
> > Consumer: Sends a readWrite flag and portletContext
> > Producer: Returns a portletContext with the *same* portletHandle
and
> > possibly *different* portletState.
> >
> > I did a quick search in the Conformance spreadsheet and
could not find a
> > statement about whether "the producer MUST return
the same
> > portletHandle along with the same/modified portletState
when consumer
> > supplies readWrite".
> >
> > If we have overlooked something, let me know.
> >
> > I think the same question applies to modifyRegistration.
> >
> > Regards,
> > Subbu
> >
> > ---------------------------------------------------------------------
> > To unsubscribe from this mail list, you must leave the
OASIS TC that
> > generates this mail. You may a link to this group
and all your TCs
> in OASIS
> > at:
> > https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
> >
> >
> > ---------------------------------------------------------------------
To
> > unsubscribe from this mail list, you must leave the OASIS
TC that
> > generates this mail. You may a link to this group and all
your TCs in
> > OASIS at:
> > https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
>
>
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail. You may a link to this group and all your
TCs in OASIS
> at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
>
>
> ---------------------------------------------------------------------
To
> unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail. You may a link to this group and all your TCs
in
> OASIS at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
---------------------------------------------------------------------
To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail. You may a link to this group and all your TCs
in OASIS
at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]