- From: Amelia A Lewis <alewis@tibco.com>
- Date: 24 Oct 2002 14:56:17 -0400
- To: www-ws-desc@frink.w3.org
Web Services Description Working Group
2002-10-24 meeting minutes
Present:
Present:
David Booth W3C
Allen Brookes Rogue Wave Software
Roberto Chinnici Sun Microsystems
Youenn Fablet Canon
Dietmar Gaertner Software AG
Steve Graham Global Grid Forum
Tom Jordahl Macromedia
Jacek Kopecky Systinet
Sandeep Kumar Cisco Systems
Philippe Le Hégaret W3C
Amelia Lewis TIBCO
Steve Lind AT&T
Kevin Canyang Liu SAP
Jonathan Marsh Chair (Microsoft)
Dale Moberg Cyclone Commerce
Jean-Jacques Moreau Canon
Don Mullen Tibco
Jeffrey Schlimmer Microsoft
Igor Sedukhin Computer Associates
William Stumbo Xerox
Jerry Thrasher Lexmark
Steve Tuecke Global Grid Forum
William Vambenepe Hewlett-Packard
Sanjiva Weerawarana IBM
Don Wright Lexmark
Joyce Yang Oracle
Prasad Yendluri webMethods, Inc.
Regrets:
Glen Daniels Macromedia
Arthur Ryman IBM
Waqar Sadiq Electronic Data Systems
-----------------------------------------------------------------------
Agenda
http://lists.w3.org/Archives/Public/www-ws-desc/2002Oct/0110.html
1. Assign scribe
2. Approval of minutes of Oct 17 telcon
3. Review of action items
4. Administrivia
5. Requirements
6. New issues
7. Importing schema into WSDL
8. Identification of a service / Arthur's identification proposal
9. A slice at a proposal for SOAP features/properties in WSDL
10. Rationale for dropping the soap:body/@use attribute.
11. BindingType proposal from Kevin
12. Issue 2: SoapAction has been deprecated, as of SOAP 1.2
13. Issue 28: soap:binding/@transport attribute
14. Issues 6a, 41: HTTP binding
15. Issue 25: Interaction between W3C XML Schema and SOAP data model
-----------------------------------------------------------------------
1. Assign scribe
Amy Lewis (TIBCO/Extensibility)
-----------------------------------------------------------------------
2. Approval of minutes of 17 Oct telcon
Approved
-----------------------------------------------------------------------
3. Review of action items
DROP 2002-09-09: Sanjiva to redo part 3.2 of his binding proposal
PENDING 2002-09-10: Sanjiva to prodice a proposal for equivalence of
top-level components.
? 2002-09-10: Gudge, jeffsch, roberto et al to write proposal
to replace message with complexType
? 2002-09-10: Gudge to provide summary of using W3C XML Schema
to wrap other type systems at an appropriate level
of abstraction.
? 2002-09-11: Jeffrey, DonM to define TCP binding.
SKIP 2002-09-19: Joyce, Sandeep, Igor, Steve T, Sanjiva, Adi,
Roberto, Amy to form task force to prepare
presentation on adding pub/sub as first class to
WSDL 1.2.
PENDING 2002-09-19: Sanjiva to write a Java binding.
DONE? 2002-10-10: Gudge to roll port type inheritance into the main
branch, adding an issue that we need to resolve the
name issue and also the issue for promoting
operations to the top level.
PENDING 2002-10-10: Gudge to prepare proposal for promoting operations
to the top level.
? 2002-10-10: Roberto to see how this might affect the feature
proposal. Glen to review.
DONE 2002-10-17: JSchlimmer to finish requirements document.
DONE 2002-10-17: JMarsh to investigate the first item listed as
new issue, since the link in the agenda is broken.
DONE 2002-10-17: ALewis to provide feedback on list for feature
proposal.
PENDING 2002-10-17: Gudge, ALewis to make proposal for language
describing schema extensibility, and for using
xs:import/xs:schema as examples of this.
-----------------------------------------------------------------------
4. Administrivia
Reminder of DST ending.
Pallavi Malu (Intel) has left the group.
Philippe: apologies for spam on list, problem resolved.
Last call for W3C process document, please review.
-----------------------------------------------------------------------
5. Requirements
JMarsh: What status is requirements document to have? Suggest last
call. [somebody] requests clarification of requirement 120.
ACTION: JMarsh to provide document ready for publication next week,
to be voted on by WG.
-----------------------------------------------------------------------
6. New issues
6a. Attempt to clarify language on schema extensibility.
Jacek: objects that language of 2.3.2 is inconsistent with goal of
schema extensibility. ALewis: language included in current mods,
submitted to Gudge. Wait for Gudge review, publication.
6b. service/@serviceType not defined in schema.
JMarsh: moot issue, to be removed. JMarsh: should we try harder to
synch schema/spec? General agreement that synch failure happens prior
to publication. Not an issue.
6c. Extension attributes cannot be required.
Does this need a highlight in the spec? General agreement; marked
editorial.
6d. Open content model and ability of extensions to specify parentage.
Gudge suggests substitution groups (on list). ALewis suggests further
investigation before adopting this solution, integration with features
proposal. Roberto notes that substitution groups do not work for
attributes. Conclusion: investigate further.
6e. Do we need an issues editor?
Responsibility: maintain document tracking issues opened, closed,
etc. Sanjiva seconds, others agree. Jacek notes successful use in XMLP
WG. Call for volunteers; Sanjiva volunteers. Accepted
ACTION: Sanjiva to act as issues editor, create document for tracking
issues.
-----------------------------------------------------------------------
7. Importing schema into WSDL
On list, Gudge supplied explanation of xs:import vs. xs:include.
Sanjiva now opposed to addition of include. Suggest single term.
Comments: confusion with xs:import (similar syntax, different semantics)
considered bad thing (reference to Don Box on list). Roberto: prefer to
have same semantics as W3C XML Schema. Possibility of single WSDL
contianing xs:import, xs:include, and wsdl:using raised as potentially
confusing. Resolution postponed for Gudge feedback; straw poll next
week.
DMullen on disallowing embedded schemas: info on list, but short
summary is that creates issues for schema provisioning mechanisms at
large scale. Draws back from forbidding; suggests best practice.
Sanjiva objects that there are no problems with scaling. Issue of reuse
underlined. TJordahl: not user-friendly to require split of single
service definition into multiple documents. DMullen provides use case
for automated verification of schema. Various comments made on whether
general-purpose tools will support WSDL and schema extraction
therefrom. General sentiment: make no recommendations. Suggest issue
be explored in primer. Primer editor agrees.
ACTION: DBooth to discuss scenarios for embed/import of schemas in
primer.
-----------------------------------------------------------------------
8. Identification of a service / Arthur's identification proposal
Initially: defer discussion of Arthur's proposal (not present).
DBooth: suggests issue relates to recently-opened issue of identifying
service semantics (see list).
Igor's proposal for clarification of service identification.
Discussion, leading to agreement that Arthur's proposal identifies the
components of a service, including the service component itself.
Igor, Sanjiva: QNames are still valid means of service
identification. Jacek: why restrict identification to one format only?
RDF identified as driver for service identification via URIs;
requirement 120 mentioned. JMarsh: does this need to be in the main
specification? JSchlimmer: problem should not be resolved for service
as a special case; solve general problem of identification of
components. Already have WSDL-internal QName mechanism; required to
solve issue for RDF mapping using URIs. Igor: most identification
strategies covered by Arthur's proposal.
Discussion of proper location for listing of identification
methodologies. QName needed for WSDL-internal, used throughout spec.
URI-based identification not needed by core spec. JMarsh suggests
location in RDF mapping document, where it can be normative for the
mapping. DBooth notes that while RDF requires this style, other
specifications may also require it. Sanjiva suggests that location of
text should not be a major issue for discussion.
JMarsh: awkward that identification sometimes requires not only QName,
and not only QName + parent, but sometimes parent and grandparent.
TJordahl agrees. Promotion of operation suggested as solution to part
of the problem. Single symbol space for top-level components also
suggested. Sanjiva objects to placing all components in single
symbol-space, not thrilled at combining symbol-spaces of all top level
components. KLiu agrees, mentions multi-part WSDL as use case. JMarsh
objects that guaranteeing uniqueness not that much harder, and part of
application responsibility. Mention made that XSD WG has recognized the
separate symbol spaces (model for WSD) as a problem, is addressing
problem with NuN mechanism.
Sanjiva, JMarsh: does Arthur's proposal do more than is necessary?
TJordahl: if approved, what is it good for? TJordahl: proposed format
provides identification only, no location. Usual fruitless argument
about URI and URL and location semantics, ending in agreement that
location is useful. JMarsh: can get location semantic by combining URL
of WSDL with greater part of Arthur's proposal. Phillippe: URI-based
identification very useful for things like choreography, other specs
built on WSDL.
ACTION: Jacek to provide commentary/response to Arthur's proposal.
-----------------------------------------------------------------------
9 - 15 (see above)
These items were not discussed.
--
Amelia A. Lewis
Architect, TIBCO/Extensibility, Inc.
alewis@tibco.com
Received on Thursday, 24 October 2002 14:56:19 UTC