Ian,
This sounds very good and I agree. Thanks for the write up.
One question of clarification. This proposal makes no changes to the
WSDL extension part of the spec and the
/wsdl:portType/@wsrmd:metadataDescriptor remains a QName. Correct?
On 4 Apr 2006, at 12:26, Ian Robinson wrote:
>
>
>
>
> On the TC call on Mon Apr 3 we discussed an alternative resolution to
> that
> originally proposed for issue 174 for obtaining the values of metadata
> descriptor document instances that might be specific to a WS-Resource
> instance and potentially dynamic (hence factored as a RP). Instead of
> introducing an additonal RP that contains the value of an RMD instance
> doc
> for a WS-Resource, this proposal changes the type and some semantics
> of the
> existing MetadataDescriptorRef RP (which contains a reference to an RMD
> instance doc).
>
> Regardless of which solution we follow, we need to start by fixing a
> problem in theRMD wd-08 draft so that:
>
> Line s 897/898,
> <xsd:element name="MetadataDescriptor"
> type="wsrmd:MetadataDescriptorRef"/>
> should be
> <xsd:element name="MetadataDescriptorRef"
> type="wsrmd:MetadataDescriptorRefType"/>
>
> This resolution then proposes the further change that the definition of
> MetadataDescriptorRefType be changed to wsa:EndpointReferenceType. The
> value of MetadataDescriptorRef is then an EPR to a "metadata
> WS-Resource"
> associated with the original WS-Resource.
>
> The description of this element should be changed as follosw:
>
> "
> The constraints on this element are as follows:
>
> /wsrmd:MetadataDescriptorRef
> This element is an EndpointReference to a "metadata WS-Resource"
> associated
> with the target WS-Resource. This metadata WS-Resource has a resource
> properties document that is the MetadataDescriptor document of the
> target
> WS-Resource. The metadata WS-Resource MUST support the
> wsrf-rpw:GetResourcePropertiesDocument message exchange and MUST
> respond to
> such a request by returning the MetadataDescriptor document that
> describes
> the target WS-Resource.
> "
> The description of
> "/wsrmd:MetadataDescriptorRef/@metadataDescriptorLocation" should be
> removed from this section.
>
>
> <TC discussion not intended to be included in the resolution>
>
> In general, if a WS-Resource "A" has an associated metadata WS-Resource
> "mdA" then the resource properties document associated with mdA *is*
> the
> metadescriptor document associated with A. Information about A's
> (potentially dynamic) instance metadata can then be queried (and
> potentially updated) using normal WSRF-RP message exchanges targeted
> at the
> mdA EPR.
> A constraint on metadata WS-Resources, above and beyond that of general
> WS-Resources, is that they MUST support the GetResourcePropertyDocument
> message exchange (which is optional for general WS-Resources). The
> constraint needs to be stated in the RMD doc in the section "10.2 Using
> Resoure Property Elements to expose MetadataDescriptors "
>
> </TC discussion not intended to be included in the resolution>
>
>
> The schema should then define:
> <!--
> <MetadataDescriptorRef>
> <wsa:Address>URI</wsa:Address>
> <tns:SomeReferenceParameter>xsd:any<tns:SomeReferenceParameter> ?
> </MetadataDescriptorRef>
> -->
> <xsd:complexType name="MetadataDescriptorRefType" >
> <xsd:complexContent>
> <xsd:extension base="wsa:EndpointReferenceType">
> </xsd:extension>
> </xsd:complexContent>
> </xsd:complexType>
>
> <xsd:element name="MetadataDescriptorRef"
> type="wsrmd:MetadataDescriptorRefType" />
>
>
>
>
>
> Regards,
> Ian Robinson
>
>
--
Take care:
Dr. David Snelling < David . Snelling . UK . Fujitsu . com >
Fujitsu Laboratories of Europe
Hayes Park Central
Hayes End Road
Hayes, Middlesex UB4 8FE
+44-208-606-4649 (Office)
+44-208-606-4539 (Fax)
+44-7768-807526 (Mobile)
