W3C home > Mailing lists > Public > www-ws-desc@w3.org > December 2003

RE: Proposal: abstract faults

From: <paul.downey@bt.com>
Date: Tue, 23 Dec 2003 19:07:51 -0000
Message-ID: <2B7789AAED12954AAD214AEAC13ACCEF0FFF1E7B@i2km02-ukbr.domain1.systemhost.net>
To: <sanjiva@watson.ibm.com>, <gdaniels@sonicsoftware.com>, <alewis@tibco.com>
Cc: <www-ws-desc@w3.org>
Sanjiva
 
I think you're right on both counts, so:
 
  <binding>
      <fault name="qname">
        <wssoap:fault faultcode="mntoken"? />
      </fault>
  </binding>*
 
Paul

	 
	-----Original Message----- 
	From: Sanjiva Weerawarana [mailto:sanjiva@watson.ibm.com] 
	Sent: Sat 20/12/2003 03:46 
	To: Glen Daniels; Amelia A Lewis; Downey,PS,Paul,XSJ67A C 
	Cc: www-ws-desc@w3.org 
	Subject: Re: Proposal: abstract faults
	
	

	+1 .. operations have QNames to enable inheritance and so should
	faults.
	
	Isn't the binding syntax a bit messed up? I think the fault
	name should be on /binding/fault rather than /binding/fault/
	wsoap:fault IIRC.
	
	Sanjiva.
	
	----- Original Message -----
	From: "Glen Daniels" <gdaniels@sonicsoftware.com>
	To: "Amelia A Lewis" <alewis@tibco.com>; <paul.downey@bt.com>
	Cc: <www-ws-desc@w3.org>
	Sent: Friday, December 19, 2003 4:57 AM
	Subject: Re: Proposal: abstract faults
	
	
	>
	> +1 to Amy - QNames would be better.
	>
	> --Glen
	>
	> ----- Original Message -----
	> From: "Amelia A Lewis" <alewis@tibco.com>
	> To: <paul.downey@bt.com>
	> Cc: <www-ws-desc@w3.org>
	> Sent: Thursday, December 18, 2003 12:44 PM
	> Subject: Re: Proposal: abstract faults
	>
	>
	> >
	> > On Thu, 18 Dec 2003 17:37:32 +0000
	> > paul.downey@bt.com wrote:
	> >
	> > > TBH I'd prefer to avoid QNames if at all possible. I thought as there
	> > > was only one interface in a WSDL 2.0, an NCName was sufficient.
	> >
	> > Huh?  Interface inheritance means that, in WSDL 2.0, you could have
	> > lots&lots (that's more than "many", I think) of interfaces in a single
	> > document.  And lots&lots&lots more once you start importing and
	> > including.
	> >
	> > > *but* for orthogonality the fault name should be of the same type as
	> > > operation name in the <binding>. Looking at the <binding>, i notice
	> > > the operation name is linked to the interface using a QName.
	> > >
	> > > Does that mean that a binding can refer to an operation in another
	> > > WSDL ?
	> >
	> > In an imported or included WSDL, you mean?  Yes.  Note that import
	> > requires a different namespace than the definitions/@targetNamespace of
	> > the current WSDL.
	> >
	> > Amy!
	> > --
	> > Amelia A. Lewis
	> > Architect, TIBCO/Extensibility, Inc.
	> > alewis@tibco.com
	> >
	> >
	
	

Received on Tuesday, 23 December 2003 14:09:52 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:27 GMT