W3C home > Mailing lists > Public > www-ws-desc@w3.org > January 2004

Part 1, section 2.4.1, {message}

From: Bijan Parsia <bparsia@isr.umd.edu>
Date: Wed, 28 Jan 2004 14:26:14 -0500
Message-Id: <D63F24CB-51C7-11D8-B510-0003939E0B44@isr.umd.edu>
To: www-ws-desc@w3.org

I *think* this is in discharge of (one of) my action item(s).

"""A Message Reference component associates XML element declarations 
that define the message content for one of the messages participating 
in an operation."""

I understand XML element declarations defining message content as 
contraining the message content to match the element declaration. Is 
this correct? Does this rule out simple types? I really don't 
understand how this can sensible permit other type systems. E.g., I 
don't see, contra what Jacek has told me, how to define the message 
content for a message to be an OWL Class.

"""The purpose of a Message Reference component is to associate an 
actual message type (XML element declaration for message content) for 
the message that will perform a specific role in the message exchange 
pattern."""

This basically defines the "actual" message type with an "XML element 
declaration for message content". Now these seem to be some strange 
term of art, but in so far as this is the same as the above, I think I 
object.

"""{message} A reference to an XML element declaration. This element 
represents the content or "payload" of the message."""

I'm trying to read this liberally, so that the element is in the 
component model only, but it doesn't seem right. The constraint seems 
to be that the content *is* the element. At the least, I need a better 
explication of "represents" in that sentence.

So, a terminology point:
	@messageReference actually identifies a "position" (for want of a 
better term) in a message exchange pattern. It doesn't reference any 
message.

	@message does not have as its content a, well, message. It identifies 
a message *type*, so I'd like to call it message type. I don't like 
element because I want to associate non-elements with the type of a 
message.

 From the RDF mapping perspective, using OWL Full for the moment, you'd 
expect to have assertions along the lines of:
	wsdl:operation wsdl:hasInput _:x.
	_:x wsdl:message my:PersonClass.
	my:PersonClass rdf:type owl:Class.

Whereas, I *think* a strict adherence to the current language would 
require something like
	wsdl:operation wsdl:hasInput _:x.
	_:x wsdl:message ''<<<an xml element definition>>"^^XMLLiteral.

I really only what the latter when the input's type is actually xml 
(from the interface perspective).

I suspect I'm still totally confused on this point. I've not been able 
to successfully explain this issue to any of the various communities I 
am a member of who might care about this issue. So if there is some 
straightening out the Working Group could provide, I would be 
exceedingly grateful.

Cheers,
Bijan Parsia.
Received on Wednesday, 28 January 2004 14:26:13 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:28 GMT