Re: Issue 225: accommodating non-XML data models (proposal)

Mark Nottingham wrote:
> On Jun 16, 2004, at 12:57 PM, Roberto Chinnici wrote:
>>  I'm less enthusiastic about
>> renaming the wsdl:types element because users are familiar with it
>> from WSDL 1.1; nevertheless, we should at least highlight in the spec
>> the fundamental infoset-centric assumption that we're making.
> 
> Hmm, that doesn't seem to have stopped the WG WRT interface, etc. Do 
> others have thoughts here?
> 
> FWIW, the "elements" attribute and its relationship to the "types" 
> element has always been one of the more non-intuitive things in WSDL to 
> me, and I'd very much like to correct this, if we can. Telling people 
> who are familiar with WSDL 1.1 that "types" is now "elements" doesn't 
> seem like a great stretch, at all.

I find "types" more natural because that's what we use them for, to
type the input and output messages of operations. The fact that
we use "element declarations" is a technical detail mostly due to
the way XML Schema is defined. (At one time, we were considering using
only model groups even!) On the other hand, if a majority of the WG
prefers "elements" (or even "elementDeclarations", if we really want to
spell out in full what goes in that section), I could live with that.

> Also, do you object to adjusting the sentence re: data model so that 
> it's specific to the types section (my #1)?

No, I don't. How about this:

   Type system components are declarations drawn from some type system.
   In particular, global element declarations as defined in XML Schema
   are type system components. They define the [local name], [namespace
   name], [children] and [attributes] properties of an element
   information item. Declarations drawn from other type systems are
   allowed. If they define information equivalent to that of a XML Schema
   global element declaration then they can be treated as if they were
   such a declaration for the purposes of this specification. Otherwise,
   their use in a WSDL document will require the use of specially-defined
   constructs to be introduced via extensibility.

Roberto

Received on Wednesday, 16 June 2004 17:01:44 UTC