RE: Semantics of WSDL vs. semantics of service

Dear Dr. McDermott,

What and how machine can do is based on what and how human designs the
procedure pre-defined for machine. Beyond the pre-defined scope, machine
cannot do anything. Like SW problem discussed in McCool's paper, machine
cannot process any exception beyond pre-defined logical inferential rules.
That's the same to SWS. So, if a human can compose something, machine can do
that based on human design. Not vice versa. 

What I emphasized is how to enable human beings in the general community to
understand the composition process, not programmers and AI professionals as
yourself. WSDL/SOAP are just for programmers and machines. That's a problem
and you should not ignore it. You cannot just imagine you and machine can do
something and that's enough, since your stress on machines obviously
aggravate the digital divide crisis. If we can find simple and easy ways to
do the same thing to replace the complex mechanism, it is your
responsibility as a scientist to share such research to the public and
community, but not on the contrary, to continue to build more complex
systems to do the same simple and easy things for the human beings. 

In http://www.mapwv.com/basemap/viewer.htm, we can integrate the maps
offered by USGS and Microsoft into this West Virginia node of National Map.
Whenever you nevigate on the map, machine can automatically "compose" a new
integrated map. How machine can do this automatically? Because I, human
being, know how to compose it first.

Best wishes,

Xuan

-----Original Message-----
From: Drew McDermott
To: public-sws-ig@w3.org
Sent: 3/16/06 11:17 PM
Subject: RE: Semantics of WSDL vs. semantics of service



> [Shi, Xuan]

> ... I know I said something different from the others. Such as what is
> service composition? My definition is different from the so-called
> "standard" meaning, but I think it is more realistic and
understandable for
> all users who are not programmers and AI professionals to consume Web
> services. 

This admission is simply astonishing.  I pointed this discrepancy out
to Xuan months ago in private correspondence.  The SW community
disagrees about many aspects of the "service composition problem," but
everyone agrees it involves computers doing some sort of combination
of solutions of small web-service problems in order to solve bigger
problems.  Because the English word "compose" is ambiguous, and
because Xuan came into this area as an outsider, he originally thought
it meant human composition of web-service requests (as one would
compose an SQL request, for instance).  An honest misunderstanding.

But it is not an acceptable response to such a revelation to continue
to use the semi-standard term in one's nonstandard way.  To do so is
to guarantee that any discussion using the term will be meaningless,
chaotic, and ultimately acrimonious.  (The more so if there are
_other_ terms that are being used in nonstandard ways; who knows?)

I don't see why anyone would pursue this any further.

-- 

                                         -- Drew McDermott
                                            Yale University
                                            Computer Science Department

Received on Friday, 17 March 2006 15:25:12 UTC