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

RE: Why single-interface is broken

From: Jacek Kopecky <jacek@systinet.com>
Date: 02 Jun 2003 15:53:42 +0200
To: WS-Description WG <www-ws-desc@w3.org>
Cc: Jeffrey Schlimmer <jeffsch@windows.microsoft.com>, "Amelia A. Lewis" <alewis@tibco.com>
Message-Id: <1054562022.2272.32.camel@localhost>

Hi all,

It seems to me that someone has a simpler, nicer model of WS description
that they are trying to push into WSDL 1.2. I'd be much more comfortable
if it was shown in its wholeness and not introduced step-by-step by
partial simplifications which may be debatable when standing alone.
Single interface per service is one such partial simplification.

I do see one concrete new capability that the attribute targetResource
brings us - adding ports to a resource without having to redefine the
existing services. This capability doesn't need the limitation of one
interface per service, though.

I think some proponents of one interface per service state that it's
good for discovery. How about a more radical step in this direction: in
order to facilitate discovery instances of a given interface, let's get
rid of the <service> element completely and add the targetResource
attribute on <port>. Compared to the current situation, we will lose
nothing but one WSDL construct that doesn't seem to have much of a value
any more.

<tongue-in-cheek>Oh, and then we could go and gather all the ports with
the same targetResource under one new construct, say <resource> and
remove the targetResource attribute. And then rename <resource> to
<service>. 8-)</tongue-in-cheek>

Best regards,

                   Jacek Kopecky

                   Senior Architect
                   Systinet Corporation
                   http://www.systinet.com/





On Sat, 2003-05-31 at 01:50, Jeffrey Schlimmer wrote:
> > From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org]
> On
> > 
> > Redundancy is *bad*.  Specifying interface twice is *bad*.  If it's
> > going to happen at all, error-handling MUST be specified in sufficient
> > detail that two processors faced with the same description report the
> > same thing.
> 
> I don't disagree with your other points, but want to point out that
> there is a proposal on the table to eliminate the specification of an
> interface (was portType) on the binding to eliminate the redundancy
> between the service/@interface and binding/@interface.
> 
> http://lists.w3.org/Archives/Public/www-ws-desc/2003May/0046.html
> 
> --Jeff
Received on Monday, 2 June 2003 09:53:56 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:58:25 GMT