- From: Bijan Parsia <bparsia@isr.umd.edu>
- Date: Mon, 26 Sep 2005 11:11:42 -0400
- To: "Asir Vedamuthu" <asirveda@microsoft.com>
- Cc: "Jacek Kopecky" <jacek.kopecky@deri.org>, "WS-Description WG" <www-ws-desc@w3.org>
On Sep 26, 2005, at 10:24 AM, Asir Vedamuthu wrote: > Jacek, > > Thank you for posting this document. Reading through, I have few high > level questions for you: > > (a) What are the differences between this draft and three other > previous > proposals? Was it three or four? It is an refinement and completion. > (b) Was this draft produced with the help of RDF or Semantic Web > Interest Group? We presented the approach to the semantic web best practices working group at the tech plenary in boston. It had strong overlap with the Semantic Web Interest Group set of attendees (RDF interest group just *is* the SWIG now). > (c) What is your plan of action to get this reviewed and approved when > there are only two WSDL WG members from the Semantic Web domain? Good question. My time has been, obviously, very constraint > (d) Are there any implementations or product plans that the WG can use > to demonstrate experience during the CR phase? For the process of developing the mapping tables, a person in my organization is also developing XSLT and XQuery scripts (suitable for using with GRDDL) for the transformation. > (e) Is this a 'lossy' mapping from WSDL Component Model to RDF? Yes. Well, mostly. It uses names and modeling techniques that are more congenial to people working with the rdf (I had an earlier version that strove to be a transliteration of the text where possible, but this was rejected by the semantic web best practices group). But it should be round-trippable (except for schema components, perhaps). The *intention* is to be lossy. > (f) What are the use cases for this mapping? What are the use cases for > Appendix B, 'WSDL Files to RDF'? Does the Semantic Web community need > both? I can give several, but one that leaps to my mind is that the Data Access Working Group has talked about being able to do a kind of P2P query wherein one SPARQL server would have both content data (in rdf) and related service descriptions (in rdf) and one would query this server (using SPARQL) to find relevant related servers. Also for just querying the server about itself. They started a design for a service description language but were persuaded toward WSDL partly on the strength of being able to extend the mapping. These are my quick answers. Cheers, Bijan.
Received on Monday, 26 September 2005 15:16:20 UTC