W3C home > Mailing lists > Public > www-ws-desc@w3.org > February 2002

Re: Reference requirements

From: Nirmal Mukhi <nmukhi@us.ibm.com>
Date: Thu, 21 Feb 2002 10:52:50 -0500
To: Jacek Kopecky <jacek@systinet.com>
Cc: Paul Prescod <paul@prescod.net>, www-ws-desc@w3.org
Message-ID: <OF7B3C7245.E2115507-ON85256B67.0056F13F@pok.ibm.com>

Hello Jacek,

Yes, I think passing around a URI to a WSDL would be good and that's the
solution I prefer; I just wanted to distinguish this from passing around
just the URI for an endpoint (or some specialized version of it) which I
think is a limited solution (since it assumes that whoever possesses the
reference knows the binding details).


                      Jacek Kopecky                                                                                                    
                      <jacek@systinet.c        To:       Nirmal Mukhi/Watson/IBM@IBMUS                                                 
                      om>                      cc:       Paul Prescod <paul@prescod.net>, <www-ws-desc@w3.org>                         
                                               Subject:  Re: Reference requirements                                                    
                      02/21/2002 06:00                                                                                                 

 as I indicated before, we view instance IDs as a kind of
sessioning mechanism. There may be other session IDs bound with
the remote reference, too, therefore I doubt a simple URI
extension would soon become unhandy. Also, I'm not sure every URI
scheme used for addressing web services allows such extensions.
 Passing the whole WSDL around might be viewed as overkill, in
our WASP product we decided to just pass a reference to the WSDL
so that the WSDL file can then be cached in various ways.
 If the WSDL description can carry additional parameters of the
web service (like QoS parameters, instanceIDs and other session
IDs etc.), passing only a reference to a WSDL file (and the QName
of the particular service in the file) can work quite nicely.
 Best regards,

                   Jacek Kopecky

                   Senior Architect, Systinet (formerly Idoox)

On Wed, 20 Feb 2002, Nirmal Mukhi wrote:

 > Hi,
 > Returning a specialized URI for an instance sounds like an excellent
 > however, I would take it one step further: why not return a specialized
 > WSDL? That gives you flexibility in terms of allowing the instance to
 > describe a binding too - which means you could ask a "factory service"
 > an instance that offers some set of port types, and the factory would
 > back to you with a WSDL (with port type definitions omitted and possibly
 > bindng omitted too - these would be imported so it would still logically
 > a complete service description). The advantage of doing this over a URI
 > that a URI isn't self-contained - I can't send you my service URI unless
 > you know everything else (the port type, binidng details). Of course
 > exchanging WSDLs is more heavyweight (even though in most cases the WSDL
 > may not contain more than <import...> and <service...> elements), but
 > there is room for both alternatives? For my part, I think using WSDL
 > as a service reference makes more sense.
 > Nirmal.

 >                       Paul Prescod

 >                       <paul@prescod.net        To:

 >                       >                        cc:

 >                       Sent by:                 Subject:  Re: Reference
 >                       www-ws-desc-reque

 >                       st@w3.org



 >                       02/19/2002 03:07

 >                       PM



 > Anne Thomas Manes wrote:
 > >
 > > Systinet WASP supports remote references using a header element to
 > reference
 > > the instance id.
 > >
 > > Anne Thomas Manes
 > > CTO, Systinet
 > That's great. Can we standardize it?
 > Even better...why not combine the endpoint URI and the instance ID
 > somewhat like this:
 > http://www.manes.net/service?instance=instanceID
 > I have other ideas beyond that but I'll let that one sink in. If you do
 > that then voila you've made it easy for every instance to be an endpoint
 > and all you need is a WSDL for it. Which puts the ball back in WSDL's
 > court. You need a way to say that the return value of a method will be a
 > URI like that and declare the WSDL that goes with that
 > instance-endpoint.
 >  Paul Prescod
Received on Thursday, 21 February 2002 11:22:02 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:54:37 UTC