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

Re: Comments on the Web Services Architecture Usage Scenarios document (was Re: New draft of the Web Services Architecture Usage Scenarios document)

From: Mark Baker <distobj@acm.org>
Date: Thu, 20 Jun 2002 22:45:07 -0400
To: www-ws-arch@w3.org
Message-ID: <20020620224507.B28158@www.markbaker.ca>

On Thu, Jun 20, 2002 at 06:51:59PM -0400, Hugo Haas wrote:
> S600 Address based Discovery[2] is useful and relates to the "no a
> priori knowledge" requirement. However, the description proposed is
> complex: it seems to me that if somebody has a URL for something (a
> service here), the natural thing to do is to do a GET on it.

I want to emphasize the importance of this approach.  Of course, I
don't believe that the URI of the Web service should respond with the
WSDL on a GET, since GET is supposed to return a representation of
the state of the resource/service.  But a related URI that identifies
the interface would be fine.

So if we have this URI that identifies the Web service;


then a GET on it could return in the header (SOAP or HTTP),
something equivalent to;

Interface; /wsdl

which is an assertion (I suppose I'm not allowed to use RDF, sigh) that
would signal to the client that it could do a GET on;


to retrieve the WSDL.

> Under candidate technologies for S600, I would list HTTP GET and the
> output of the Web Services Description Working Group. It seems to me
> that WSIL[3] is more a catalog of services and their descriptions,
> which is a different kind of usage scenario IMO. BTW, I was looking at
> the specification, and was wondering how generic and how Web
> service-specific it was. I was wondering if something like RDDL[4]
> could be used in such a way.

IMO, WS-Inspection exists because most Web services developers don't
take advantage of the a priori agreement than you have with HTTP.  So
WS-Inspection attempts to solve that problem by introducing a layer of
indirection to eventually get you to a place where GET can happen, when
it could have been done on the Web service URI in the first place!

This issue also relates to S601; a candidate technology for a registry
would be any hypermedia document containing one or more URI.

BTW, my previous comments on this document still hold;


Mark Baker, CTO, Idokorro Mobile (formerly Planetfred)
Ottawa, Ontario, CANADA.               distobj@acm.org
http://www.markbaker.ca        http://www.idokorro.com
Received on Thursday, 20 June 2002 22:34:54 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:05:34 UTC