RE: Comments on WSDL 1.1 Element Identifiers

I considered that case for longer than I'd care to admit, but in the end
felt that it flew in the face of import as we understand it.  That is, if I
have already imported a particular namespace, I can ignore any other imports
(location is just a hint, right?)  Similarly, if a WSDL declares a
targetNamespace, wouldn't a subsequent attempt to import that namespace
simply be a no-op? 

The modularity example in the WSDL 1.1 spec uses different namespaces for
each of the files (schema, portType, service/binding).  So while the spec
doesn't explicitly make importing the same namespace illegal, I would be
surprised if anybody were abusing import in this way (and even more
surprised to find that any kind of interoperability resulted.)

Jonathan Marsh - http://www.wso2.com - http://auburnmarshes.spaces.live.com
 

> -----Original Message-----
> From: Amelia A. Lewis [mailto:alewis@tibco.com]
> Sent: Tuesday, February 13, 2007 1:22 PM
> To: Jonathan Marsh
> Cc: www-ws-desc@w3.org
> Subject: Re: Comments on WSDL 1.1 Element Identifiers
> 
> One comment:
> 
> Jonathan Marsh wrote:
> > My claim that imports and includes make designating a WSDL element
> difficult
> > are false to because is no include in WSDL 1.1, and WSDL 1.1 imports
> require
> > a namespace.  There will thus be a 1-1 correspondence between a WSDL 1.1
> > document and a particular target namespace.
> 
> Err, I don't believe that this is the case.
> 
> Specifically, it is possible (and in fact an example is given in the
> WSDL 1.1 specification) to "partition" a WSDL into pieces, all of which
> have the same namespace.  In other words, a WSDL 1.1 document can import
> another WSDL 1.1 document that has the same namespace as the importing
> document, and the practice is more or less recommended in the WSDL 1.1
> specification as a means of modularizing definitions.
> 
> WSDL 1.1 does not, as I understand it, specify behavior in cases in
> which an import redefines an existing definition, if those definitions
> differ.
> 
> Consequently, while it may be possible to create ambiguous pointers due
> to the multi-purpose import/include/dessert-wax nature of WSDL 1.1
> import, that ambiguity is going to reflect ambiguity within the set of
> documents.  Arguably, broken pointers are better suited for broken
> documents than working ones.  Right?  :-)
> 
> Amy!
> --
> Amelia A. Lewis
> Senior Architect
> TIBCO/Extensibility, Inc.
> alewis@tibco.com

Received on Tuesday, 13 February 2007 22:29:14 UTC