W3C home > Mailing lists > Public > www-ws-desc@w3.org > June 2004

RE: Issue 130: Asynch request/response HTTP binding needed

From: <paul.downey@bt.com>
Date: Thu, 24 Jun 2004 16:06:58 +0100
Message-ID: <2B7789AAED12954AAD214AEAC13ACCEF0FFF2366@i2km02-ukbr.domain1.systemhost.net>
To: <sanjiva@watson.ibm.com>, <tomj@macromedia.com>, <jmarsh@microsoft.com>, <www-ws-desc@w3.org>

all or some of this may be related to Issue 130.:

Expectation was very high (at least within my company) that 
WSDL 2.0 would say *something* regarding asynchronous Web service 
interactions. This is something we have great need in describing and
have to roll our own solutions ATM in the absence of an open and 
interoperable standard.

So if you can bear with me i've some use-cases to work through 
before going down the "say nothing" path:

  1) a MEP is implemented on a transport which
     provides a return path for a receiver back to a 
     sender (e.g. a HTTP connection, a 'replyTo' address).

  2) a MEP is implemented using a transport which doesn't 
     provide a return path for a receiver to communicate 
     with the sender, or the return path isn't back to 
     the original sender.

  3) one or more separate operations are used to implement a 
     higher level MEP, such as in the WS-I callback scenario[1]

Case (1) is OK with WSDL 2.0 so long as the bound transport 
(sync or async) has a return address mechanism.

Case (2) requires an addressing mechanism to be carried in the 
message contents - WS-Addressing, WS-MessageDelivery or some such.

Unfortunately we don't have a standard we can normatively point
to, and as Dave says, this is akin to dispatching: we don't want 
to possibly pick the wrong horse or preclude other future mechanisms 
- so it's best kicked out to extensions for WSDL 2.0.

Issue 130 AIUI directly relates to (3). Dave's proposal[2] seemed 
to address this issue, but i think in a hard-wired fashion? 
The two endpoints being described in WSDL up front and tied together. 
Whilst this could be useful, i'd imagine most people would prefer 
to use a ReplyTo field carried dynamically in the message. 
This puts (2) and (3) into the same bucket IMO.

So, sadly, "say nothing" sounds like the most pragmatic route for 
this WG, and all i have to do now is work out how to break the bad 
news to my colleagues.

Paul

[1] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jun/0217.html
[2] http://lists.w3.org/Archives/Public/www-ws-desc/2004May/0029.html


-----Original Message-----
From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org]On
Behalf Of Sanjiva Weerawarana
Sent: 24 June 2004 02:30
To: Tom Jordahl; 'Jonathan Marsh'; 'Web Services Description'
Subject: Re: Issue 130: Asynch request/response HTTP binding needed



Same here; there is nothing called an "asynch" pattern IMO. As
you Jonathan noted nothing precludes one from doing In-Out with
asynch stuff .. in fact the use of WS-Addressing ReplyTo, for
example, already allows that.

Sanjiva.
Received on Thursday, 24 June 2004 11:07:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:31 GMT