W3C home > Mailing lists > Public > www-ws-desc@w3.org > July 2004

RE: Part 1: Component reference vs. QName

From: Martin Gudgin <mgudgin@microsoft.com>
Date: Mon, 19 Jul 2004 18:35:22 -0700
Message-ID: <DD35CC66F54D8248B6E04232892B633802C3C601@RED-MSG-43.redmond.corp.microsoft.com>
To: "Roberto Chinnici" <Roberto.Chinnici@Sun.COM>, "Sanjiva Weerawarana" <sanjiva@watson.ibm.com>
Cc: "Asir Vedamuthu" <asirv@webmethods.com>, <paul.downey@bt.com>, <www-ws-desc@w3.org>

I've not looked at the spec in detail recently but my recollection is
that at the component model level we tried to avoid having properties
which were references, favouring properties which were actual components
instead. Such properties were populated by following references at the
syntax level.

Gudge

> -----Original Message-----
> From: www-ws-desc-request@w3.org 
> [mailto:www-ws-desc-request@w3.org] On Behalf Of Roberto Chinnici
> Sent: 19 July 2004 07:33
> To: Sanjiva Weerawarana
> Cc: Asir Vedamuthu; paul.downey@bt.com; www-ws-desc@w3.org
> Subject: Re: Part 1: Component reference vs. QName
> 
> 
> We're not proposing to change the syntax, so you'd still be using
> a @ref attribute of type xsd:QName to refer to operations/faults from
> a binding.
> 
> The proposal is to change the component model so that the 
> {operation reference}
> property of the Binding Operation and the {fault reference} 
> property of
> Binding Fault would have the actual components as a value 
> instead of a QName.
> The motivation being that, because of the rules already in 
> place, the current
> QNames resolve uniquely to a component in _all_ cases, 
> inheritance or not.
> 
> Roberto
> 
> 
> Sanjiva Weerawarana wrote:
> > Right, but spse
> > 
> >     interface x:I1 extends y:I2 [xmlns:x=foo1, xmlns:y=foo2]
> > 
> > then the binding which specifies a binding for x:I1 must also
> > bind the operations/faults in y:I2 .. which have a different TNS.
> > 
> > Sanjiva.
> > 
> > ----- Original Message ----- 
> > From: "Roberto Chinnici" <Roberto.Chinnici@Sun.COM>
> > To: "Sanjiva Weerawarana" <sanjiva@watson.ibm.com>
> > Cc: "Asir Vedamuthu" <asirv@webmethods.com>; <paul.downey@bt.com>;
> > <www-ws-desc@w3.org>
> > Sent: Monday, July 19, 2004 8:36 PM
> > Subject: Re: Part 1: Component reference vs. QName
> > 
> > 
> > 
> >>How? In the context of an interface, operations and faults 
> are uniquely
> >>identified by a qname, so qname-typed references to operation/fault
> >>components and the components themselves are interchangeable.
> >>
> >>Roberto
> >>
> >>
> >>Sanjiva Weerawarana wrote:
> >>
> >>>Didn't you forget our wonderful inheritance model???
> >>>
> >>>Sanjiva.
> >>>
> >>>----- Original Message ----- 
> >>>From: "Roberto Chinnici" <Roberto.Chinnici@Sun.COM>
> >>>To: "Asir Vedamuthu" <asirv@webmethods.com>
> >>>Cc: <paul.downey@bt.com>; <www-ws-desc@w3.org>
> >>>Sent: Saturday, July 17, 2004 4:17 AM
> >>>Subject: Re: Part 1: Component reference vs. QName
> >>>
> >>>
> >>>
> >>>
> >>>>Asir,
> >>>>
> >>>>I was in the process of writing a thorough explanation of 
> why the spec
> >>>>is the way it is, but I think you ran into a real issue.
> >>>>
> >>>>At some point we allowed "generic" binding components 
> (i.e. those with
> >>>>an unspecified {interface} property) to contain Binding Fault and
> >>>>Binding Operation components, but that functionality is 
> gone (see third
> >>>>paragraph of section 2.9.1). The use of QNames to refer 
> to Interface
> >>>>Fault/Operation(s) from in Binding Fault/Operation(s) is 
> a vestige of
> >>>>those days.
> >>>>
> >>>>As things stand now, this use of QNames is inconsistent 
> with the rest
> >>>>of the specification, so I'd be in favor of using actual 
> components
> >>>>instead.
> >>>>
> >>>>Unless I'm missing something, of course!
> >>>>
> >>>>Roberto
> >>>>
> >>>>
> >>>>Asir Vedamuthu wrote:
> >>>>
> >>>>
> >>>>>Paul,
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>>>There was some discussion of this following the proposal to
> >>>>>>hoist faults:
> >>>>>>http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/
> 0062.html
> >>>>>
> >>>>>
> >>>>>Thank you. I read this thread. My question is not at the XML
> >>>
> >>>representation
> >>>
> >>>
> >>>>>level but, at the -component- level. Let me quote from part 1,
> >>>>>
> >>>>>"{fault reference} REQUIRED. A wsdls:QName as defined by 
> 2.15.5 QName
> >>>
> >>>Type
> >>>
> >>>
> >>>>>which refers to an Interface Fault component in the 
> {faults} property
> > 
> > of
> > 
> >>>the
> >>>
> >>>
> >>>>>Interface component identified by the {interface} property of the
> > 
> > parent
> > 
> >>>>>Binding component. This is the Interface Fault component 
> for which
> >>>
> >>>binding
> >>>
> >>>
> >>>>>information is being specified." [1]
> >>>>>
> >>>>>Let me re-state my question. {fault reference} property 
> appears to be a
> >>>>>component reference. Per part 1, {fault reference} property is a
> >>>>>wsdls:QName.  Thus, the following two properties stand out,
> >>>>>
> >>>>>(a) Binding Fault Component.{fault reference}
> >>>>>(b) Binding Operation Component.{operation reference}
> >>>>>
> >>>>>Their values are of type wsdls:QName instead of Interface
> >>>
> >>>Fault/Operation
> >>>
> >>>
> >>>>>component. Is that intentional?
> >>>>>
> >>>>>[1]
> >>>>>
> >>>
> >>>
> > 
> http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl20/wsdl20
> .html?content-
> > 
> >>>>>type=text/html%3B0charset=utf-8#Binding_Fault_details
> >>>>>
> >>>>>Asir
> >>>>>
> >>>>>-----Original Message-----
> >>>>>From: paul.downey@bt.com [mailto:paul.downey@bt.com]
> >>>>>Sent: Thursday, July 15, 2004 9:41 AM
> >>>>>To: asirv@webmethods.com; www-ws-desc@w3.org
> >>>>>Subject: RE: Part 1: Component reference vs. QName
> >>>>>
> >>>>>
> >>>>>Asir,
> >>>>>
> >>>>>AIUI fault and operations are identified using ncnames, 
> but referenced
> >>>>>using qnames, since the same fault name may exist in one or more
> >>>
> >>>interface.
> >>>
> >>>
> >>>>>There was some discussion of this following the proposal to hoist
> >>>
> >>>faults:
> >>>
> >>>
> >>>>>http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0062.html
> >>>>>
> >>>>>HTH
> >>>>>Paul
> >>>>>
> >>>>>
> >>>>>
> >>>>>-----Original Message-----
> >>>>>From: www-ws-desc-request@w3.org 
> [mailto:www-ws-desc-request@w3.org]On
> >>>>>Behalf Of Asir Vedamuthu
> >>>>>Sent: 15 July 2004 13:39
> >>>>>To: www-ws-desc@w3.org
> >>>>>Subject: Part 1: Component reference vs. QName
> >>>>>
> >>>>>
> >>>>>
> >>>>>In part 1 component model, the following properties (see 
> below) appear
> >>>
> >>>to be
> >>>
> >>>
> >>>>>component references but, they are described as QNames. Is that
> >>>
> >>>intentional?
> >>>
> >>>
> >>>>>- Binding Fault Component.{fault reference} [1]
> >>>>>- Binding Operation Component.{operation reference} [2]
> >>>>>
> >>>>>[1]
> >>>>>
> >>>
> >>>
> > 
> http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl20/wsdl20
> .html?content-
> > 
> >>>>>type=text/html%3B0charset=utf-8#Binding_Fault_details
> >>>>>
> >>>>>[2]
> >>>>>
> >>>
> >>>
> > 
> http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl20/wsdl20
> .html?content-
> > 
> >>>>>type=text/html%3B0charset=utf-8#Binding_Operation_details
> >>>>>
> >>>>>Regards,
> >>>>>Asir S Vedamuthu
> >>>>>asirv at webmethods dot com
> >>>>>http://www.webmethods.com/
> 
> 
Received on Monday, 19 July 2004 21:35:40 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:32 GMT