W3C home > Mailing lists > Public > www-ws-arch@w3.org > July 2002

Re: REST, Conversations and Reliability

From: Paul Prescod <paul@prescod.net>
Date: Sat, 20 Jul 2002 17:34:05 -0700
Message-ID: <3D3A017D.28A146A3@prescod.net>
To: "David W. Levine" <dwl@watson.ibm.com>, www-ws-arch@w3.org
CC: "Champion, Mike" <Mike.Champion@SoftwareAG-USA.com>

"David W. Levine" wrote:
> 
>...
> 
> As one final example, it was pointed out that returning WSDL as a response
> to an HTTP get on a
> SOAP URI is undesirable because it doesn't directly map to REST. The
> underlying question is
> in what specific ways is this undesirable, and how can the desired effect
> be achieved.

David, you said a lot of things worth listening to. I'll just address
this question because I think I know at least one answer.

Even if you ignore REST, returning WSDL from GET is a little strange
because WSDL is something you need at design time. Pointing to it at
runtime is at best a little convenience, not a core feature. We should
not "waste" the runtime GET on a design-time feature. It is way too
valuable for actually retrieving a representation of the resource's
state. Even if the resource is a "big component" as in older SOAP
services, it could return its runtime state and a *pointer* to the WSDL.
It is easy for the runtime representation to point to the static design
time object but harder for the static WSDL to point to the current
runtime state. This is especially true in a universe where many
resources use the same WSDL (which is what we should be aiming for REST
or not).

And even where it is possible, indirecting through the static WSDL is
quite an inefficient way to get to the current dynamic state.
-- 
Come discuss XML and REST web services at:
  Open Source Conference: July 22-26, 2002, conferences.oreillynet.com
  Extreme Markup: Aug 4-9, 2002,  www.extrememarkup.com/extreme/
Received on Saturday, 20 July 2002 20:35:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 3 July 2007 12:25:02 GMT