W3C home > Mailing lists > Public > public-ws-addressing@w3.org > January 2006

RE: proposed breaking change to echo global element declaration

From: Simon Fell <sfell@salesforce.com>
Date: Wed, 4 Jan 2006 10:48:34 -0800
Message-ID: <D4A7A018B7CBC343B9CADCA70B44203C7EECF6@exsfo-mb02.internal.salesforce.com>
To: tim@mindreef.com, paul.downey@bt.com, distobj@acm.org, public-ws-addressing-tests@w3.org, public-ws-addressing@w3.org

I should have been more carefull in the wording of my question (Mark got
what I was driving at I think)

I understand what wsa:action indentifies, but if the tools don't support
dispatching from it, and the tests are being altered to reflect that
fact, then as a designer of a service that might want to use WS-A, it
seems like I should be real careful not to have any depenedencies on
wsa:Action, as the current tools don't fully support it, nor will the
WS-A tests indentify which tools do and do not support it. In which
case, why leave it in the spec, its just going to cause an interop
problem, or are we expecting a WS-I profile spec to come out to address
how to actually use WS-A in the real world ?

Cheers
Simon

-----Original Message-----
From: public-ws-addressing-request@w3.org
[mailto:public-ws-addressing-request@w3.org] On Behalf Of Tim Ewald
Sent: Wednesday, January 04, 2006 7:03 AM
To: Simon Fell; paul.downey@bt.com; distobj@acm.org;
public-ws-addressing-tests@w3.org; public-ws-addressing@w3.org
Subject: RE: proposed breaking change to echo global element declaration


It conveys the intent of the message without having to crack the body
(esp.
if the body is encrypted). I haven't read the latest draft of the spec,
but I raised an issue with an earlier draft around mandating that
wsa:Action be used for dispatching. Among other things, it means that a
kit has to dispatch differently when using WS-Addressing and when not
using WS-Addressing. But more importantly, why does should the spec care
how dispatching is done?

Thanks,
Tim-

> -----Original Message-----
> From: public-ws-addressing-request@w3.org
> [mailto:public-ws-addressing-request@w3.org] On Behalf Of Simon Fell
> Sent: Tuesday, January 03, 2006 1:26 PM
> To: paul.downey@bt.com; distobj@acm.org; 
> public-ws-addressing-tests@w3.org; public-ws-addressing@w3.org
> Subject: RE: proposed breaking change to echo global element 
> declaration
> 
> 
> If dispatching is done purely on the GED, then what exactly is 
> wsa:Action for ??
> 
> Tx
> Simon
> 
> -----Original Message-----
> From: public-ws-addressing-request@w3.org
> [mailto:public-ws-addressing-request@w3.org] On Behalf Of 
> paul.downey@bt.com
> Sent: Tuesday, January 03, 2006 9:46 AM
> To: distobj@acm.org; public-ws-addressing-tests@w3.org;
> public-ws-addressing@w3.org
> Subject: RE: proposed breaking change to echo global element 
> declaration
> 
> 
> Hi Mark,
> 
> > Whatever the resolution, I hope that the need for additional 
> > out-of-band agreement (other than what is defined in the SOAP & WSDL
> > specs) is documented as an interoperability problem.
> 
> I'm unsure what OOB agreement is needed - the issue is that the same 
> element 'echo' being used for the input and output, an endpoint acting

> as both the sender and receiver roles can only use Action to 
> distinguish if the echo is a request or a response.
> 
> In practice, an endpoint is unlikely be operating as both 'client'
> and 'server', 'publisher' and 'subscriber' etc. so changing the 
> message body element to be also unique relaxes the requirement 
> artificially placed upon implementations to process based upon 
> wsa:Action rather than the message contents or even some combination 
> of the two.
> 
> As far as the sender is concerned, they only need to populate the 
> wsa:Action and body fields as directed by the WSDL, EPR or whatever 
> and all is well.
> 
> Paul
> 
> 
> 
Received on Wednesday, 4 January 2006 18:48:53 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:35:11 GMT