- From: Jacek Kopecky <jacek.kopecky@systinet.com>
- Date: Mon, 13 Oct 2003 18:23:10 +0200
- To: Bijan Parsia <bparsia@isr.umd.edu>
- Cc: www-archive@w3.org
Bijan, sorry it took me so long to get back to you with this. In my review of Appendix E [1] I noted that if we decide to limit WSDL messages (not currently defined anywhere AFAIK, but consisting of body and headers *in the form of XML Element declarations*) only to the body and headers, we may require the use of XML Schema to describe the inputs and outputs of services that don't necessarily communicate using XML-Schema-describable XML, like a web service that returns a JPEG of the current view on Prague. Nobody on the call had any objections to this direction, and some suggested that for such media types as JPEG it may be appropriate to describe the data passed around using the media typing approaches currently hovering between WS-Desc WG and XMLP WG (see [2]). If we imagine a service that accepts or rejects RDF graphs, we might want to be able to describe the allowable data using RDFS/OWL. That's not describing the services, just the input/output message formats. But since the current notion of message only allows XML Element declarations, we could have to resort to using the generic rdf:RDF element and possibly state the OWL/RDFS description using extensibility. So basically - do you mind? 8-) I can see it both ways, messages being restricted to XML element declarations or messages extensible to just about any type system and data model (MIME-typed binary objects, SOAP Data Model graph nodes and edges, RDF ontologies etc.) Hope this message clarifies your action a bit, Jacek [1] http://lists.w3.org/Archives/Public/www-ws-desc/2003Sep/0136.html [2] http://lists.w3.org/Archives/Public/www-ws-desc/2003Sep/0088.html
Received on Monday, 13 October 2003 12:23:20 UTC