RE: D-AC009.2 discussion points and proposal(s)

David,
 I favor very strongly the coupling of the ws activity to the semantic web
activity for all the reasons that have been given below and for those we
have discussed in the ws.
 --Katia

-----Original Message-----
From: www-ws-arch-request@w3.org [mailto:www-ws-arch-request@w3.org]On
Behalf Of David Booth
Sent: Wednesday, May 22, 2002 6:18 PM
To: www-ws-arch@w3.org
Cc: David Orchard
Subject: RE: D-AC009.2 discussion points and proposal(s)


At 03:33 PM 5/21/2002 -0700, "David Orchard" <dorchard@bea.com> wrote:
>The W3C membership CLEARLY indicated that web services IS decoupled from
the
>Semantic Web Activity.  . . . .

Although more people spoke out in favor of decoupling Web Services work
from the Semantic Web activity than those who spoke out against such
decoupling, I think it's only fair to point out that there were strong
voices on both sides of the question.

A search of the w3c-ac-forum email list (
http://www.w3.org/Search/Mail/Team/advanced_search?keywords=%22semantic+web%
22&hdr-1-name=subject&hdr-1-query=%22web+services%22&hdr-2-name=from&hdr-2-q
uery=&hdr-3-name=message-id&hdr-3-query=&resultsperpage=100&sortby=date&inde
x-grp=Team%2FFULL+Member%2FFULL+Public%2FFULL&index-type=t&type-index=w3c-ac
-forum&index=ac-discussion
) yielded 20 hits, of which 18 (excerpted below) seem to pertain to this
issue.  Of those, it looks to me like about 10 of those who wrote were in
favor of decoupling and about 5 were against decoupling.

Here are excerpts of the search results.  I have tried to ensure that these
excerpts accurately characterize the sense of the authors' complete
messages (relative to this issue), but please refer to the original
messages for the full context if there is a question.

                       ---

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0199.html :
Tim Clark, Millennium Pharmaceuticals:
"Relationship with Semantic Web activities is potentially very important,
especially for Life Sciences applications where ontologies play a
significant role. The charter should include a statement requiring active
liaison between the Web Services activity and both W3C WebOnt and DAML
Services (DAML-S) groups."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0198.html :
Ora Lassila, Nokia:
"Relationship with Semantic Web activities is important; the current
proposal mentions this only in passing and does not sufficiently emphasize
the importance. The charter should include a statement about liaison
between the Web Services activity and both W3C WebOnt and DAML Services
(DAML-S) groups."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0186.html :
David Orchard, BEA:
"Web Services should be loosely coupled to Semantic Web and RDF. Semantic
Web suggested changes to Web Services should be evaluated by the working
groups and architecture groups, but not mandated in any charters."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0076.html :
Don Deutsch, Oracle:
'we do NOT favor mandating the use of the Resource Description Framework
(RDF) for "any semantically significant information" at this time.'

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0071.html :
Eric Newcomer, Iona:
"The semantic web represents a wonderful vision, but Web services represent
the next significant practical use of the web. We should ensure the success
of Web services first, and the top priority. The semantic web effort stands
to be more meaningful in the context of a Web services enabled Web since
the Web will be much more useful to business and society than it is today.
. . . .   Can anyone really imagine trying to reinvent [SOAP, WSDL, and
other associated technologies] using RDF? Certainly RDF has its place, and
would have equal place in the Web services enabled world."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0070.html :
Tex Texin, Progress Software:
"RDF and Web Services should not be coupled at this time."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0060.html :
Joe Meadows, Boeing:
"We . . . believe that the Semantic Web activity is very important, and
that ignoring the ability to leverage semantically significant information
would be a major drawback to any implementation, thus we strongly encourage
that the issue not be sidestepped, but rather, be addressed head on."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0027.html :
Johan Hjelm, Ericsson:
"web services come first and then, later, semantic web technologies are
applied to the web services registries and the like.  . . . Keeping the
architecture as clean as possible, keeping dependencies as few and small as
possible, should be a goal in itself. And if dependencies exist, they need
to be documented early and it should be possible to provide a modular
implementation - preferrably one where the dependencies can be removed (so
that you can implement web services without RDF, but if you do it with RDF,
it becomes much better than otherwise... carrot instead of stick. That
would be interesting to see)."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001OctDec/0002.html :
Frederick Hirsch, Zolera Systems:
"Dynamically assembling comprehensive web services to meet specified
criteria such as quality of service or cost will require processing as
exemplified in the semantic web vision. Such XML services will require meta
information and decisions. For this reason we believe the W3C activity
should focus on the XML Service Description working group proposal and
attempt to create a new, simple yet elegant solution that fits the
architecture of the semantic web. . . .  Despite the benefits of WSDL,
considering a different semantic web based approach, and consolidating
layers, might produce a simpler, generic solution.  Rather than rushing to
approve an industry proposal, we believe the W3C should focus on producing
a simple, powerful and long term solution."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0214.html :
Alexander Falk, Altova:
"Web Services . . . are a top priority for standardization. . . . The
Semantic Web, RDF, and ontologies . . . [are] more a "research" matter,
whereas things like Web Services are a clear "development" matter and more
important in the short-term. . . .  [For] the foregoing reasons, Web
Services and RDF or the Semantic Web should be decoupled."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0211.html :
David Orchard, BEA:
"BEA Systems strongly opposes the coupling of RDF and the Semantic web to
the Web Services standardization efforts.  RDF and Semantic Web activities
may eventually provide very valuable solutions to problems facing
developers and organizations in the future. But that does not appear to be
the case today."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0193.html :
Mark Wood, Eastman Kodak:
"The work of the Semantic Web activity is not yet mature nor widely
accepted. Consequently, at this time we see no reason to mandate the use of
RDF for Web Services, although that may be appropriate in the future."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0186.html :
Michael Wilson Chair, CLRC Rutherford Appleton Laboratory:
"CLRC views the service descriptions as requiring graph beyond hierarchy
structure and therefore RDF would be more appropriate than XML Schema."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0185.html
Oisin Hurley, Iona:
"While the Web Services and Semantic Web activities may both live quite
happily as isolated works, I think there is a potential great benefit in
studying and effecting their intersection. However I do not think that this
benefit will be immediately obvious or executable, so I would say that both
activities should come to a certain level of maturity before a initiative
is undertaken to find cross-applications."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0184.html :
Renato Iannella, IPR Systems:
"IPR Systems would strongly prefer to see any proposed Activity for a Web
services language to be based on XML Schema. XML Schema should be
considered by W3C as the core schema language for "common infrastructure".
(Unfortunately, RDF/RDF Schema poses too many unknowns into the equation.)"

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0182.html :
Eve Maler, Sun:
"we're concerned about taking on small pieces of work without developing
them against a coherent vision. To this end, we believe the TAG should be
formed as quickly as possible and should immediately begin developing an
architectural vision with which the web services description work can later
align as necessary. The TAG would be the appropriate forum to consider
fully the potential relationship of the semantic web and web services."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0176.html
Roger Cutler, Chevron:
"On the face of it, the WSDL activity seems much more attractive than the
RDF-specific proposal. For one thing, it has the backing of industry
heavyweights already and considerable de facto acceptance. If the W3C goes
in some other direction I think there is a definite risk of being ignored.
Moreover, if that happened WSDL would not get the needed "working over" in
terms of integration and validation, and possibly extension, that a W3C
working group would give it, so de facto acceptance might lead to flawed
implementation. Finally, it seems to me premature to make a commitment to
using RDF, the new kid on the block, for a high priority main-line function
like this."

http://lists.w3.org/Archives/Member/w3c-ac-forum/2001JulSep/0166.html :
Franz Fritz, SAP:
"We do not see the absolute necessity to align WSDL with RDF in the first
step."



--
David Booth
W3C Fellow / Hewlett-Packard
Telephone: +1.617.253.1273

Received on Wednesday, 22 May 2002 20:29:02 UTC