Singleton pattern in WSRF (was: RE: [wsdm] Discovery Scenarios [fred])

["Sedukhin, Igor S" <Igor.Sedukhin@ca.com>]

Tom,  

I have no doubts that you are right about normativeness, but one has to read between the words in order to undersstand how to apply this. If the very first sentence of the section that describes the service group (section 5) starts from [A ServiceGroup is a WS-Resource, following the implied resource pattern]. And then, the whole document is written with this assumption in mind, it may be hard to understand what to do otherwise. Moreover, it may not be clear that if one implements a ServiceGroup in a "singleton" mode, this approach is legitimate according to the existing spec.

All I'm saying is that I'd suggest WSDM TC to recommend to WSRF TC that
1) it must be clear how to use "singleton" (regular Web services) pattern with WS-ServiceGroup, WS-ResourceProperties, etc.
2) it is clarified how to determine that WS-Resource or "singleton" patterns are used given a WSDL document
3) the "singleton" pattern is defined and allows message exchanges without ANY required headers

I believe that there is a lot of value in being able to use WS-ResourceProperties, WS-ServiceGroup, etc. in regular Web service implementations. Those operations and schemas do not have to be reinvented. It should be possible to use them with or without headers and initial EPRs.

PS. I think fulfilling 1)-3) is technically very easy.

-- Igor Sedukhin .. (igor.sedukhin@ca.com)
-- (631) 342-4325 .. 1 CA Plaza, Islandia, NY 11788

-----Original Message-----
From: Tom Maguire [mailto:tmaguire@us.ibm.com] 
Sent: Friday, May 21, 2004 12:29 PM
To: Sedukhin, Igor S
Cc: fred.carter@amberpoint.com; Wsdm (E-mail)
Subject: RE: [wsdm] Discovery Scenarios [fred]





I am pretty sure that a WS-ServiceGroup is not normatively described as having to be a WS-Resource.