OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

wsdm message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [wsdm] RE: [wsn] Re: [wsdm] Proposed requirement from WSDM


On the WSDM conf call I confirmed with the group that WSDM is ok to defer to WSN on whether the resulting notification message should comply with the schema or not.
 
William


From: Vambenepe, William N
Sent: Friday, September 10, 2004 1:01 PM
To: Steve Graham
Cc: Ugo Corda; wsdm@lists.oasis-open.org; wsn@lists.oasis-open.org
Subject: [wsdm] RE: [wsn] Re: [wsdm] Proposed requirement from WSDM

I think the heart of the requirement is "let the subscriber tell the producer what parts of the message the consumer wants". The part about "must comply to the schema" is secondary. My guess is WSDM would be happy to defer to WSN to decide on this.
 
William


From: Steve Graham [mailto:sggraham@us.ibm.com]
Sent: Friday, September 10, 2004 5:04 AM
To: Vambenepe, William N
Cc: Ugo Corda; wsdm@lists.oasis-open.org; wsn@lists.oasis-open.org
Subject: RE: [wsn] Re: [wsdm] Proposed requirement from WSDM


"Vambenepe, William N" <vbp@hp.com> wrote on 09/09/2004 07:35:01 PM:

> Yes, this is a hard part. On solution is to drop the "must comply to the schema"
> requirement, for example by saying that this subscribe option can only be used on
> topics that contain "xsd:Any" in their associated "messageType" attribute.

How important is the "must comply to schema" constraint in your scenario. Given that
the subscriber is imposing this message subset constraint, it strikes me that
the subscriber/consumer should then be capable of dealing with an arbitrary
message type that results from the subset. Quite frankly I don't see why the
NotificationProducer would care whether the message (post subsetting) would validate
or not.  To me, the validation constraint is one of producer-consumer coordination.
If the subscriber/consumer specifically chooses to have a subsetting constraint, then
the shape of the resulting message should be the subscriber/consumer's concern alone.

>  
> Another option is to say that this allows you to drop elements that have a
> minOccurs of zero. Maybe to keep it simple we also say that these elements must be
> direct children of the root. Or maybe we let them be anywhere.

The more "special casing" we do, the harder the implementation and the more cases where
the feature is not usable.  I would prefer dropping the validation constraint.

>  
> We can try to come up with solutions to this. But the first order of business is
> for this group to decide whether to open an issue on this requirement from WSDM.

Agreed.

>  
> Regards,
>  
> William
>
> From: Ugo Corda [mailto:UCorda@SeeBeyond.com]
> Sent: Thursday, September 09, 2004 11:29 AM
> To: Steve Graham; Vambenepe, William N
> Cc: wsdm@lists.oasis-open.org; wsn@lists.oasis-open.org
> Subject: RE: [wsn] Re: [wsdm] Proposed requirement from WSDM

> > It will, of course be very tricky to make sure that the "pattern" expressed will
> end up with a notification message that is

> > still validatable with respect to the message schema.
>
> I had the same type of reaction when reading the requirement. How do you guys in
> WSDM make a server accept messages that do not comply with the WSDL schema
> specified for the complete message? Do you also define submessage schemas in the WSDL?

>  
> Thank you,
> Ugo
> -----Original Message-----
> From: Steve Graham [mailto:sggraham@us.ibm.com]
> Sent: Thursday, September 09, 2004 11:08 AM
> To: Vambenepe, William N
> Cc: wsdm@lists.oasis-open.org; wsn@lists.oasis-open.org
> Subject: [wsn] Re: [wsdm] Proposed requirement from WSDM

>
> Interesting.
> So this is kinda like a very strongly applied filter expression?
> Right now, the selector specifies a pattern that MUST be met in order for the
> entire message to be sent.
>
> If I understand this requirement, it is an additional "expression" on the
> subscribe, but stating the pattern of the notification itself, ie the elements that must
> appear on the notification and no more than those elements specified.
>
> It will, of course be very tricky to make sure that the "pattern" expressed will
> end up with a notification message that is still validatable with respect
> to the message schema.
>
> sgg
>
> ++++++++
> Steve Graham
> (919)254-0615 (T/L 444)
> STSM, On Demand Architecture
> Member, IBM Academy of Technology
> <Soli Deo Gloria/>
> ++++++++
>
>

>
> "Vambenepe, William N" <vbp@hp.com>

> 09/09/2004 01:34 PM
>
> To

>
> <wsn@lists.oasis-open.org>

>
> cc

>
> <wsdm@lists.oasis-open.org>

>
> Subject

>
> [wsdm] Proposed requirement from WSDM

>
>
>
>
> Dear WSNers,
>
> The WSDM TC would like to submit the following requirement to WSN:
>
> "Add the option, when subscribing for notification, to specify that only
> parts of the message should be sent. Note that the message that is sent
> still needs to be schema-valid for the schema of the message."
>
> The reasoning behind this requirement is that the content of the
> notification message can get very large and in some cases only a subset
> of this is of interest to the consumer. And in some cases, the
> notification producer doesn't know what parts of the notification the
> consumer is interested in, but the subscriber knows. So it should be
> possible for the subscriber to express this in the subscription.
>
> The WSDM TC would like WSN to add this to the WSN issues list.
>
> Regards,
>
> William
>
> PS: doesn't it make us all feel good that Outlook wants to correct
> "WSNers" to "Winners" on the first line of this email? ;-)
>
> To unsubscribe from this mailing list (and be removed from the roster of the OASIS
> TC), go to http://www.oasis-open.org/apps/org/workgroup/wsdm/members/leave_workgroup.php.
>


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]