This is the complete version, possibly with confidential material, of the minutes.
1. Assign scribe. Lucky minute taker for this week is: Steve Lind
2. Approval of minutes: the minutes [.1] of the previous conference call were approved
[.1] http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0066.html
3. Review of Action items [.1].
? 2003-03-13:
Don will write a proposal for annotating schema
with part information.
? 2003-03-27:
Philippe write up a proposal for embedding binary
data types in schema
? 2003-05-13: DaveO to send a motivating example for R131.
? 2003-05-13:
Jeffsch, Sanjiva, Glen, Umit, JJM to come up with
a proposal to get rid with the message construct,
and add programming hints.
? 2003-05-14:
Kevin to contact Sanjiva and try to merge
proposals.
? 2003-06-12:
Jacek to synthesize the different approaches
to solving issue 64.
DONE [.2] 2003-06-12: Arthur to summarize his concerns
about Umit's
proposal by email.
DONE [.4] 2003-06-19: Roberto to rewrite up an email about
the whether
to use "xs:any" or substitution groups, or both,
for our extensibility mechanism.
DONE [.3] 2003-06-19: Igor to write up about pros and
cons of the
single service proposal.
DONE [.3] 2003-06-19: Igor to come up with a new diagram.
[.1] http://www.w3.org/2002/ws/desc/#actions
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0142.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0127.html
[.4] http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0147.html
4. Administrivia
a. July FTF logistics [.1], registration [.2]
<sdl-scribe> JM: f2f: start wed morning; end early on Fridayb. 16 May 2003: The Web Services Task Force [.3] of the
<sdl-scribe> JM: good idea to review Internationalization WG draft and see if any ideasc. XPath, XQuery, XSLT Last Call document review [.5].
<sdl-scribe> JM: Is there any reason or interest to review XSL WG documents?[.1] http://www.w3.org/2002/ws/desc/3/05/f2fJulyLogistics.htm
<sdl-scribe> David: assume people would be interested based on desire to use, not a direct correlation with Desc WG
<sdl-scribe> JM: we're not on the hook for any formal response
<sdl-scribe> ... Any volunteers to review?
* jeffsch believes XPath etc is v. important but regrets to say that he's overcommitted at present...
<sdl-scribe> David: anyone looking to use XPath?
<sdl-scribe> Arthur: we should probably use the most recent spec
<sdl-scribe> JM: Don't want to get into the subset game again
<sdl-scribe> Arthur: 1.0 a subset of 2.0? is there backward compatibility?
<sdl-scribe> JM: for the most part
<sdl-scribe> JM: as a WG, we won't do it and no one is volunteering
5. Task Force Status.
a. Properties and Features (dormant)
b. Patterns
<sdl-scribe> doing survey of what people think of variations
<dbooth> http://lists.w3.org/Archives/Public/public-ws-desc-meps/2003Jun/0029.html
<sdl-scribe> David: variations of same basic pattern we've always been discussing
c. Attributes
d. QA & Testing
6. New Issues. Merged issues list [.1].
- Error in diagram (David) [.2]. Jacek's
additional (?) problem [.3].
<sdl-scribe> David: diagram shows interfaces pointing to resource; not what we currently have in data model
<sdl-scribe> ... we have service pointing to resource
<sdl-scribe> Arthur: purpose not to show syntax
<sdl-scribe> David: but should show symantics
<sdl-scribe> Arthur: resource directly implements the interface
<sdl-scribe> ... interface is reusable
<sdl-scribe> Umit: can have same interface related to two different resources
<dbooth> the _resource_ does NOT implement the interface. The _service_ implements the interface.
<sdl-scribe> Jonathan: would it help not to have an arrow there?
<sdl-scribe> ... Description of the interface can be reused
<sdl-scribe> JacekK: if we called the thing an instance of the interface; bindings between that and the resource
<sdl-scribe> David: need to be precise in our terminology
<sdl-scribe> ... WSDL construction of service is an abstract thing; interface syntax describes an abstract thing we can call an interface
<Philippe> Jacek's diagram: http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0117.html
<sdl-scribe> JacekK: Interface is exposed by its URL
<sdl-scribe> David: given a resource isn't necessarily a piece of software, how can it have an interface?
* jeffsch notes that the term "resource" is subtle and that smart people have a difficult time agreeing on its precise definition
<sdl-scribe> Arthur: David trying to turn diagram into close representation of the WSDL
<sdl-scribe> Resource is black box; don't know how it is built but can do things with it.
<sdl-scribe> can group actions into an interface
* alewis notes that "resource" seems to be used as a weasel-word, having approximately the meaning "wotsit"
<sdl-scribe> bindings and end point are still more concrete
<sdl-scribe> binding specifies the mapping of the concrete end point
<sdl-scribe> JacekK: disagree - interface can be applied to many different resource or resource can have many different interfaces
<sdl-scribe> for each interface, can have different bindings
<sdl-scribe> Jonathan: all things in diagram are concrete?
* JacekK agrees that what Jonathan just described is the source of confusion - different terms 'interface' and 'binding' in the diagram and in WSDL
<sdl-scribe> Arthur: Resource exposes the interface
<sanjiva> +1 to not having arrows!
<sdl-scribe> ... need to say what the arrow means
* sanjiva didn't realize the picture graph had a circularity
* Ingo agrees with umit
<sdl-scribe> Jonathan: have 2 camps - directly aligned with syntax or descibes an instance
<dbooth> ACTION: dbooth to propose definitions for "interface", "service", etc. as they pertain to the diagram
* RRSAgent records action 1
* alewis applauds JacekK
<dbooth> +1 to Jacek's comments
<jeffsch> +1 that removing wsdl:service is an odd (at best) suggestion
<sanjiva> +1
* jeffsch wonders if the diagram is actually helping the WG and/or readers of the specification...
* umit agrees with jeffsch (for a change :-))
* jeffsch ... are we trying to precisely define something that doesn't materially affect how the description language works?
<sdl-scribe> Jonathan: scrap the diagram?
* sanjiva thinks that the picture provides a clean semantic .. but if people can't get it and/or agree with one picture its hopeless
<sdl-scribe> David: sweeping it under the rug is not going to help peoples understanding
<Arthur> pls post the url of Jacek's ascii art
<Marsh> http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0117.html
<Ingo> http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0117.html
<sdl-scribe> Jonathan: do we need to include the targetResource given the thread on renaming?
<dbooth> "equivalence" of endpoints will _always_ be application-defined -- not defined by the WSDL language itself, no matter what the WSDL spec says.
<sdl-scribe> Umit: need to fix the diagram or remove it
<sdl-scribe> Jonathan: keep in WSDL the capability to tell which endpoints have equivalence
<Arthur> proposed definition of equivalence of endpoints is that WSDL should provide a means for providers of a service to indicate to clients of that service that they are free to select any endpoint from a set of given endpoints
<sanjiva> <service name="xx"> <alternative-endpoints><endpoint ../> <endpoint ../></alternative-endpoints> .. </service>
<sanjiva> (I think that's what Jeff is suggesting.)
<Arthur> furthermore, there is no implied formal definition of equivalence
<Arthur> there is only a claim by the provider that the client may choose any endpoint from a specified set
<umit> +1 to Arthur. All we are trying to do is to indicate to a client the possible multiple ways to access the same thing.
<Arthur> why not use <choice> :-)
<sanjiva> hey, let's allow people to use XSD syntax .. then we can have choices, sequences, all or whatever you want!
<sdl-scribe> JacekK proposing dropping targetResource attribute
<sdl-scribe> JeffSch proposing to drop the diagram as well
<sdl-scribe> Jonathan: target resource has value from a discovery standpoint
<sdl-scribe> objection to removing targetResource?
<sdl-scribe> JeffSch: may be people not on the call that will object
<Arthur> so targetNamespace == targetResource !
<sanjiva> Jack: We dropped definitions/@name a while ago.
* sanjiva remembers that we have to put interfaces in different TNSs already for potential inheritability
<jeffsch> Proposal to remove targetResource and leave feature of indicating 'equivalence'
<jeffsch> ... between endpoints as out of scope.
<jeffsch> Propose removing diagram and related text in Part 1.
<sdl-scribe> AmY: uncomfortable restricting service to a sinle interface
* dbooth thinks if you can't add things to your description then you've written it wrong. You should have used "include" or "imports".
<sdl-scribe> Is picture describing service or resource?
<sanjiva> To me the picture was attempting to provide semantics to a Web service. The circle showing what <wsdl:service> doesn't provide any semantics.
<sanjiva> JM point 1: anyone against removing @targetResource
<sdl-scribe> no one on call seems to object
* jeffsch except W3C\David as a proxy
<sanjiva> Q1: Should WSDL retain the information whether different endpoints are related or not?
<sanjiva> Glenn: @ F2F we decided yes
<sanjiva> Jack: we would reject the idea of indicating relations b'ween services until someone comes up with a definition of that relation and then we'd reconsider.<sdl-scribe> JM: out of time, can't reach decision
* sdl-scribe thanks all who contributed to the minutes
[.1]
http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/issues/wsd-issues.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0064.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2003Jun/0117.html
Scribe: Steve Lind