- From: FABLET Youenn <fablet@crf.canon.fr>
- Date: Mon, 30 Sep 2002 09:23:45 +0200
- To: www-ws-desc@w3.org, jmarsh@microsoft.com
- Message-ID: <3D97FC01.805@crf.canon.fr>
Web Services Description Working Group
2002-09-26 conference call minutes.
Attendance
Present:
David Booth W3C
Allen Brookes Rogue Wave Software
Glen Daniels Macromedia
Youenn Fablet Canon
Dietmar Gaertner Software AG
Martin Gudgin Microsoft
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
Arthur Ryman IBM
Waqar Sadiq Electronic Data Systems
Jeffrey Schlimmer Microsoft
Igor Sedukhin Computer Associates
William Stumbo Xerox
Jerry Thrasher Lexmark
Steve Tuecke Global Grid Forum
William Vambenepe Hewlett-Packa
rd
(x:47) Sanjiva Weerawarana IBM
(x:30) Don Wright Lexmark
Joyce Yang Oracle
Regrets:
Roberto Chinnici Sun Microsystems
Steve Graham Global Grid Forum
Jeff Mischkinsky Oracle
Adi Sakala IONA Technologies
Barbara Zengler DaimlerChrysler Research and Technology
-------------------------------------------------------------------
Agenda
1. Assign scribe
2. Approval of minutes of Sept 19 telcon
3. Review of Action items
4. FTF planning
5. Arch harvesting of WSDL. Need reviewers.
6. Requirements
7. New Issues
(none)
8. Rationale for Dropping the <soap:body use=...> Attribute
9. BindingType proposal from Kevin
10. A slice at a proposal for SOAP features/properties in WSDL
11. Issue 2: SOAPAction has been deprecated, as of SOAP 1.2
11. Issue 28: transport='uri' [15]
Dependent upon Glen's feature/property proposal.
12. HTTP Binding Issues (6a, 41
)
13. Issue 25: Interaction between W3C XML Schema and SOAP Data Model
-------------------------------------------------------------------
1. Assign scribe
Youenn Fablet
-------------------------------------------------------------------
2. Approval of minutes of Sept 19 telcon
Done
-------------------------------------------------------------------
3. Review of Action items
PENDING 2002-07-21: Don Mullen to write up an issue on making the
transport attribute match the SOAP binding
framework.
PENDING 2002-09-09: Sanjiva to redo part 3.2 of his binding proposal.
PENDING 2002-09-09: Gudge to check whether there is already an issue
against Part 2: can you define different
encodingStyles for different children of
soap:Body (message parts).
IN PROGRESS 2002-09-10: Steve, William, Arthur, Gudge, Joyce to produce a
proposal for portType extensibility. Reword: Steve
and Gudge to write up the portType extensibility
proposal.
PENDING 2002-09-10: Sanjiva to produce a proposal for equivalence of
(at least) top-level components in the next couple
of weeks.
PENDING 2002-09-10: Gudge; jeffsch; roberto et al to write proposal
[to remove message and replace with complexType.]
IN PROGRESS 2002-09-10: Gudge to provide summary of using xml schema to
wrap other type systems at an appropriate level
of abstraction.
PENDING 2002-09-11: Sanjiva to describe out/out-in for pub-sub. [I
think this should be pub-sub _without_ out/out-in.]
PENDING 2002-09-11: Jeffrey and Don define TCP binding.
PENDING 2002-09-19: Joyce, Sandeep, Igor, Steve T, Sanjiva, Ad
i,
Roberto, Amy to form a task force to prepare
presenation about adding pub/sub in a first
class manner to WSDL 1.2.
PENDING 2002-09-19: Sanjiva to write a Java binding.
PENDING 2002-09-19: Glen to draft SOAP last call comment on why
SOAPAction is not a "feature", and request the
ability to set arbitrary mime headers.
New Actions as of 2002-09-26
Philippe to get Eric's comments
Jeffrey to remove the word draft for req125
Jonathan to prepublish the req docs
Jacek to make a proposal for better describing the extensibility mechanism to support other languages
-------------------------------------------------------------------
4. FTF planning
Jonathan: please register to the next F2F
no progress on january F2F
people think that we should meet at the tech plenary. Will make progress on that
-------------------------------------------------------------------
5. Arch harvesting of WSDL. Need reviewers.
Jonathan: Any reviewer? My answer is: it looks good
Arthur: What is it for? A short summary of WSDL?
Jonathan: Yes
???: are the last two points WSDL issues?
Jonathan: Maybe, this is paul prescott stuff. Does it belong to WSDL scope?
Arthur: it is related to service references. We need a way to describe services that return references to other services
Jonathan: already an issue
Arthur: It might be the hot WSDL issue in the mind of the Arch group
-------------------------------------------------------------------
6. Requirements
Jonathan: No response from RDF people concerning the semantic web requirement (R120). Then I may be happy to drop this draft requirement
Philippe: I take an AI to get Eric's comments.
ACTION: philippe to get Eric's comments
Jonathan: I will send to XMLP and Arch group before publishing to public. Do not want to go in the publishing process twice
Jonathan: what about req125? Reject this req?
Jean-Jacques: It is included in SOAP1.2. If we want to fully support it, we should accept this as a requirement
ACTION: jeffrey to remove the word draft for req125
ACTION: jonathan to prepublish the req docs
-------------------------------------------------------------------
7. New Issues
(none)
-------------------------------------------------------------------
8. Rationale for Dropping the <soap:body use=...> Attribute
Jacek: I was an opponent of dropping the use attribute. If we keep the ability to use different schema languages, I have no objection to drop it. If we drop the use attribute, we have 2 options: hack XML Schema or use different schema languages. Sanjiva says that we should keep this feature in WSDL1.2; ok for dropping the use attribute.
Jonathan: Is this the only thing to make clear in Arthur's message?
Jacek: make clear we are retaining multiple XML Schemas
Arthur: also note that messages are replaced by Schemas
Jacek: when saying remove messages by schemas, what is "schemas"? Only XML Schema?
Arthur: The question is: what is the core binding of WSDL?
WS-I recommends use=litteral and have XML Schema.
If use=encoded with SOAP encoding, put it in the SOAP binding part.
Push the use choice in the SOAP binding
Jacek: pb : XMLSchema not good for SOAP encoding structures
two ways: mapping from XML Schema to SOAP encoding or allow multiple schema languages which is clearer.
XML Schema is made for tree not for other data model. It is hacky to use XML Schema then map it to SOAP encoding. We should allow multiple schema languages at the abstract level.
Gudge: point is WS-I can describe all SOAP encoding structures they need with XML Schema
Jacek: do they show the mapping?
Gudge: no mapping. They describe graphs using XML Schema
Jacek: We have a W3C XMLP work effort. Is it possible not to use SOAP encoding with WSDL?
Philipe: Yes. XMLP WG will not be surprised
Arthur: WS-I does not say we must not use SOAP encoding. WSDL does not restrict what you put on the wire.
Steve: Note that RDF may not be described by XML Schema
Gudge: they abuse of XML syntax: five different possibilities to describe a graph
Steve: Why tying WSDL with XML Schema? Why can't we support Relax NG?
Gudge: at now, we support the use of other schema languages
Steve: I see some movement to only use XML Schema in WSDL
Arthur: the point is: what goes on the SOAP binding. Additional info might be put on the SOAP binding to solve interop problems. But I think XML Schema is sufficient
Jacek: if we are sure that the use problem is not related to the use of multiple schema languages, then ok
Jonathan: I have difficulties imagine tying WSDL with one version of XML Schema
Arthur: Just by producing a new WSDL version for a new XMLSchema version. I prefer unambiguous use of schemas
Jacek: agree on unambiguous but using other schemas is not about ambiguousity: it extends WSDL
Jonathan: WS-I profile says: use XML SChema1.0 with WSDL. This does not prevent to come up with a more general solution
Jonathan: To be agnostic with schemas, we should put qname on all type elements?
Gudge: No
Jonathan: Who can take up that work for clean up ? Arthur?
Arthur: Jacek is the main proponent
Jacek: I am on vacation next week. Agree to do that work after
ACTION: Make a proposal for better describing the extensibility mechanism to support other languages
Jacek: I am ok with Arthur's rational if we can use multiple schema languages. We should extend the rational to say that we allow the use of multiple schema languages.
-------------------------------------------------------------------
9. BindingType proposal from Kevin
Jonathan: no message this week on that subject. IS that because people are afraid of making that change before 1 portType extensibility 2 Glen's proposal 3 a more general change in the binding approach?
Jacek: I would mark this proposal at 2.0 material and delay it...
Jonathan: there was interest last week in the WG. Do people have questions for Kevin?
Sanjiva: hi folks
Kevin: I like the idea of delaying this until other questions on binding come up.
Jonathan: this should be related to which topic?
Gudge: tie it with part2 of the AM
Jonathan: have the part2 editors begun works on that?
Jeffrey: beginning
Jonathan: begin to do stuff on that and let Kevin's proposal dependent on the AM part2
Sanjiva: it has effects on part1 because it is a kind of binding framework
Jonathan: It seems to effect also how part2 AM will relate with binding data.
-------------------------------------------------------------------
10. A slice at a proposal for SO
AP features/properties in WSDL
Jonathan: Glen, can you make a brief intro?
Glen: the idea is to be able to describe features that can be implemented through bindings/modules
Jonathan: translate soap features as messaging features?
Glen: good question. Jean-Jacques sent comments to me and said it was more general.
IMO, it seems more general. It is a good opportunity to move this in the right direction.
Jean-Jacques: it is a good proposal. But we may push the envelope: we could describe protocols through features. It is not a lot of effort.
Glen: I agree
Jean-Jacques: I am ready to spend some time on this
Glen: also, there is nothing in the SOAP spec that ties features with SOAP1
Glen: the main point is to say: this concept of security in SSL is the same as the one you implement with SOAP through that mechanism.
Jonathan: a problem with example1: feature elements should have the wsdl namespace, not the soap namespace
Sanjiva: what is the meaning
of the correlation uri for instance?
Glen: behind this uri, you describe the native http request/response correlation meaning
Jeff: really interesting. I have hard time understanding the implications of the feature mechanism.
Alewis: I have been developing materials for XMPL WG like features and MEP. Describe bits as a feature increases dramatically its reusability.
Glen: In the SOAP binding Task Force, what a binding could do was a question. We did not want to restrict to HTTP or request/response protocol. Then, we create a correlation feature that HTTP does natively.
David: Is it like describing a mechanism for referencing a concept in an ontology? The ability to do that allows reusing it.
Jean-Jacques: another feature is the email correlation feature.
Jean-Jacques describes the email correlation feature
Jean-Jacques: the abstract correlation feature can be implemented via this email feature
Glen: an abstract feature can be implemented throu
gh a binding or modules. HTTP supports natively the request-response MEP. If you do not use the http protocol, the binding can do specific things to support this mep.
Jeff: In the absence of the design that the XMLP WG has done, our work might not be more general than WSDL1.1. It is very important for us to understand this work deeply. Is that fair?
Glen: yes
Jeff: We should understand it before changing the bindings
Glen: We can do the simple stuff
Jeff: is this your proposal
Glen: yes
Jean-Jacques: we need to do something like Glen's proposal to support SOAP1.2
Sandeep: does this mean that to support SOAP1.2, we need to have the SOAP binding at the abstract layer.
Glen: In a sense yes.
Gudde: we could do this with modular bindings
Jeff: then, we will need to write a normative way to describe all combinations of the matrix
Jonathan: Better for interop to say: implement this cell.
Jonathan: another thing in your proposal glen: the c
orrelation module was required but not the enc module. What is the purpose of describing features that are available but not required
Glen: It is to say that these modules are available. By picking values in the WSDL file, you can know precisely how to use a feature.
Jonathan: what is the purpose of saying that you can use an unlisted module.
Glen: On the Internet, you can send whatever message you want. We are not in a world where everything will be described in WSDL. No reason to use only described features.
Sanjiva: somebody has to know what modules to use whether described in WSDL or in a book.
Alewis: one can use available features not published?
Glen. Yes. It seems an unnecessary constraint to disallow this.
Jonathan: So, where are we going with this?
Glen: it is connected to other issues like use the same uri as in the SOAP framework
Sanjiva: I want to understand what that means to us
Glen: we must describe needed info, like webm
ethods or meps. Then the matter is to define the framework that will do that.
Jonathan: Should I put this at the CG level or is it premature?
Glen: why not?
Jonathan: no objection?
Jonathan: net step is to do homework on this topic and bring questions back to Glen and Jean-Jacques.
Jonathan: glen, you said that this proposal was a sketch. What is the next step?
Glen: describing the general pattern
<scribe>???</scribe>
Jonathan: We should also describe how this would affect the AM part1
Glen: I will try
--------------------------------------------------------------------
Call Adjourned
--------------------------------------------------------------------
Scribe: Youenn Fablet
Received on Monday, 30 September 2002 03:27:03 UTC