W3C home > Mailing lists > Public > public-ws-addressing@w3.org > November 2004

Re: i038: Scope of message addressing properties [i038]

From: Marc Hadley <Marc.Hadley@Sun.COM>
Date: Mon, 29 Nov 2004 10:35:00 -0500
To: Jacques Durand <JDurand@us.fujitsu.com>
Cc: "'Mark Nottingham'" <mark.nottingham@bea.com>, public-ws-addressing@w3.org
Message-id: <3B21DCE8-421C-11D9-A193-000A95BC8D92@Sun.COM>

On Nov 24, 2004, at 10:45 PM, Jacques Durand wrote:
>
> ReplyTo would have little value if restricted to a single MEP 
> instance: it should
> be usable to support callbacks and larger MEPs.
>
> I would even argue that the presence of ReplyTo in the request of a 
> Request-response MEP should not be construed as automatically applying 
> to the response of this MEP.
>
Sounds like ReplyTo has no real semantics in your view, its just a 
convenient holder for an application specific EPR ?

> (shouldn't a synchronous rsponse always get back to sender)
>
Apparently not.

> > property. E.g. Does the presence of a wsa:ReplyTo in the response
> > message of a request-response MEP indicate that future messages in 
> the
> > same 'logical' conversation should be sent to the EPR in the ReplyTo.
>
> Yes, but why would we have to say anything about the expected behavior 
> ("should"). The  receiver will know what to do with ReplyTo: depends 
> on the application logic.

I'm not sure I agree, I want to develop middleware that takes care of 
addressing, I don't want every application to have to handle it itself.

>  All what matters for WS-Addressing is that the sender knows when it 
> should include ReplyTo element or not (e.g. required by WSDL, or by 
> the EPR - but has to be per operation...)
>
Presumably by 'sender' you mean the sending application, but as I said 
above I want to be able to take care of WS-Addressing in the protocol 
stack rather than require each application to have to handle it. For 
that to work I need some senatics to be associated with the things like 
ReplyTo and FaultTo otherwise they just become additional application 
message payload.

Marc.

>
> -----Original Message-----
> From: Mark Nottingham [mailto:mark.nottingham@bea.com]
> Sent: Wednesday, November 24, 2004 4:36 PM
> To: Marc Hadley
> Cc: public-ws-addressing@w3.org
> Subject: Re: NEW ISSUE: Scope of message addressing properties [i038]
>
>
>
>
> This is now issue 038;
>    http://www.w3.org/2002/ws/addr/wd-issues/#i038
>
>
>
> On Nov 23, 2004, at 11:32 AM, Marc Hadley wrote:
>
> >
> > Title: Scope of message addressing properties
> >
> > Description: What is the expected scope of each message addressing
> > property. E.g. Does the presence of a wsa:ReplyTo in the response
> > message of a request-response MEP indicate that future messages in 
> the
> > same 'logical' conversation should be sent to the EPR in the ReplyTo.
> >
> > Justification: During discussion of issue 28[1] it emerged that there
> > were different views on the scope of message addressing properties.
> > Some WG members (myself included) assumed that message addressing
> > properties were scoped to a particular instance of an MEP while 
> others
> > believe that their impact can extend beyond that scope. The
> > specification doesn't appear to support any one view.
> >
> > Target: Core
> >
> > Proposal: This issue requires some discussion by the WG prior to
> > proposing a resolution
> >
> > Marc.
> >
> > [1] See thread starting at:
> > http://www.w3.org/mid/9D1B57A0-3349-11D9-AEEE-000A95BC8D92@Sun.COM
> > ---
> > Marc Hadley <marc.hadley at sun.com>
> > Web Technologies and Standards, Sun Microsystems.
> >
> >
> >
>
> -- 
> Mark Nottingham   Principal Technologist
> Office of the CTO   BEA Systems
>
---
Marc Hadley <marc.hadley at sun.com>
Web Technologies and Standards, Sun Microsystems.
Received on Monday, 29 November 2004 15:34:59 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 18:35:00 GMT