W3C home > Mailing lists > Public > www-ws-desc@w3.org > November 2003

Re: HTTP binding options

From: Mark Baker <distobj@acm.org>
Date: Thu, 13 Nov 2003 10:02:50 -0500
To: www-ws-desc@w3.org
Message-ID: <20031113100250.W2875@www.markbaker.ca>

On Thu, Nov 13, 2003 at 08:13:14PM +0600, Sanjiva Weerawarana wrote:
> 
> I'd say we support HTTP GET for idempotent operations just fine- 
> if you read what Mark said, the idea of the URI being dependent
> on the message contents is not a required part of Web architecture.
>
> Even the SOAP Response MEP doesn't go as far as (5) - that has
> a no-input requirement. So we're really in virgin territory by
> going that far and its at a pretty high price (lots of restrictions
> on the input (follow RPC style + simple types only), weirdness of
> whether the input is really the input and a pretty complex language
> to boot).

From a Web POV, #5 is really a GET based form, except as WSDL it
would typically be discovered at design time rather than at runtime.

I'm not sure if the WG wants to make the, IMO, very large leap from
service description language to form declaration language.  If it does,
I'd be happy to offer my assistance, as I've spent a lot of studying
hypermedia interactions and the role that forms play there.  FWIW, that
work lead me to produce an RDF based forms language; you should be
easily able to spot the crossover point between it and WSDL 1.1
urlReplacement;

http://www.markbaker.ca/2003/05/RDF-Forms/

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
Received on Thursday, 13 November 2003 10:00:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:27 GMT