- From: Glen Daniels <gdaniels@sonicsoftware.com>
- Date: Thu, 12 Aug 2004 09:45:28 -0400
- To: <paul.downey@bt.com>, <jacek.kopecky@deri.org>
- Cc: <www-ws-desc@w3.org>
Hi Paul, all: > unfortunately as Amy has pointed out, there is there nothing > in the status quo to prevent a service from declaring that > the dispatching mechanism is "out of band", "random" or even > "undisclosed" by simply conjuring up an appropriate URI. ...the key word here being "declaring". By saying that you HAVE to somehow declare how this works, you gain clarity (from Latin "de+clare, to make clear"). Sure, someone could define an extension which means "turn off operation-name resolution", but I for one don't plan to write support for such a thing in my product. And if my tool faults when trying to build skeletons to implement a service defined in a WSDL using such an extension, at least the user will know exactly which extension caused the problem. > the outcome of all this is that the consumer of a web service > will indeed have to treat operations that accept the same > input as a single event driven operation (for non-RPC styles) > and i think that is the architectural goal of some members of this WG. I think that if what you want is a single event driven operation, you should declare, in the WSDL, a single operation which takes and/or returns a choice. You should not, IMHO, use the WSDL model to declare a bunch of operations which you then invalidate by refusing to have any interoperable way to distinguish between them. --Glen
Received on Thursday, 12 August 2004 13:45:39 UTC