Re: A WSDL media type and the TAG

Philipe,

If we introduce a new fragment syntax for URI-reference identifiers, then
we need a WSDL media type . The interpretation of fragments is defined by
the media type.

Arthur Ryman



                                                                                                                                                    
                      Philippe Le                                                                                                                   
                      Hegaret                  To:       Mark Baker <distobj@acm.org>                                                               
                      <plh@w3.org>             cc:       Jonathan Marsh <jmarsh@microsoft.com>, David Orchard <dorchard@bea.com>,                   
                      Sent by:                  www-ws-desc@w3.org                                                                                  
                      www-ws-desc-reque        Subject:  Re: A WSDL media type and the TAG                                                          
                      st@w3.org                                                                                                                     
                                                                                                                                                    
                                                                                                                                                    
                      11/04/2002 02:16                                                                                                              
                      PM                                                                                                                            
                                                                                                                                                    
                                                                                                                                                    




On Fri, 2002-11-01 at 15:37, Mark Baker wrote:
>
> On Fri, Nov 01, 2002 at 11:47:02AM -0800, Jonathan Marsh wrote:
> > > I'm not sure what you mean by the WSD WG is checking with the TAG.
> >
> > We plan to look at TAG findings relating to this issue, is all.
>
> Okey doke, then allow me to point you in the right direction;
>
> http://www.w3.org/2001/tag/2002/0129-mime
>
> Quoting the important bit;
>
> "W3C Working Groups engaged in defining a language SHOULD arrange for
> the registration of an Internet Media Type (defined in RFC 2046
> [RFC2046]) for that language;[...]"
>
> This finding defines a new process that working groups should follow
> too, that is *substantially* different than in the past.  The XMLP WG
> will be the guinea pigs for this new approach, I believe ... unless
> they delay much longer 8-/.

I had the action item to look into the TAG findings and other documents
and find out if the WG was required to create a new mime type for WSDL
or not. As Mark found out, there is no such requirement on a WG. So I
believe the real issue here would be: Why do we need to have a new mime
type for WSDL?
I didn't come up yet with a valid use case to extend the current
application/xml definition. I believe it does what we need. Of course,
it does not resolve the multiple namespace document question but neither
do the */*+xml.
Looking at the HTTP protocol, you would need a different mime type in
one case: content type negociation. If the same URI can return a WSDL
and a DAML+S document, both are application/xml. You would need mime
types to differentiate them. Using the namespace name of the root
element will not help you asking the right version to the server or the
the cache.

Philippe

Received on Monday, 4 November 2002 17:09:52 UTC