- From: Philippe Le Hegaret <plh@w3.org>
- Date: 10 Apr 2003 14:01:45 -0400
- To: www-ws-desc@w3.org
Participants
Mike Ballantyne Electronic Data Systems
Lily Liu webMethods
Jeff Mischkinsky Oracle
Kevin Canyang Liu SAP
David Booth W3C
Allen Brookes Rogue Wave Software
Youenn Fablet Canon
Tom Jordahl Macromedia
Philippe Le Hégaret W3C, chair
Amelia Lewis TIBCO
Steve Lind AT&T
Ingo Melzer DaimlerChrysler
Dale Moberg Cyclone Commerce
Arthur Ryman IBM, scribe
Jeffrey Schlimmer Microsoft
William Vambenepe Hewlett-Packard
Prasad Yendluri webMethods, Inc.
Jacek Kopecky Systinet
Jerry Thrasher Lexmark
Martin Gudgin Microsoft
Erik Ackerman Lexmark
Jean-Jacques Moreau Canon
Steven White SeeBeyond
Umit Yalcinalp Oracle
Sanjiva Weerawarana IBM (12:05)
Regrets:
Glen Daniels Macromedia
Sandeep Kumar Cisco Systems
Dietmar Gaertner Software AG
Igor Sedukhin Computer Associates
Jonathan Marsh Microsoft
Roberto Chinnici Sun Microsystems
1 Assign scribe: Arthur Ryman
2. Approval of minutes of April 3 telcon
David: the title of the minutes is not really appropriate...
http://lists.w3.org/Archives/Public/www-ws-desc/2003Apr/att-0019/20030403.irc.log.html
Minutes approved.
3. Review of Action items [.1].
? 2003-01-21: Roberto and gudge to create a branch and work up
a binding proposal based on referencing type
systems directly from operation components.
(Umit's example, Sanjiva's example, WSDL 1.1
example, and others.)
? 2003-02-27: Sanjiva to send summary of
one-portType-per-service issue.
? 2003-03-04: Editors to discuss markup for testable assertions
in the spec and come back with a strategy.
? 2003-03-04: Jonathan to recruit a QA contact for the WG.
? 2003-03-04: Jonathan to recruit a test contact for the WG.
PENDING 2003-03-13: Editors will find part 2 issues to dispatch
easily next telcon.
DONE [.5] 2003-03-13: Philippe will make a proposal concerning issue
28 using features and properties.
DONE [.2] 2003-03-13: GlenD to propose a Property representation of
SOAP action parameter of application/soap+xml
type
? 2003-03-13: Don will write a proposal for annotating schema
with part information.
? 2003-03-27: Jonathan will follow-up with editors to figure
out how to improve the prose of the spec to be
aligned with schema.
PENDING 2003-03-27: Philippe write up a proposal for embedding bindary
data types in schema
PENDING 2003-04-03: Arthur to bring discussion to group in two weeks
(more or less) for solutions to R085
PENDING 2003-04-03: Arthur to coordinate work on WSDL validator
Arthur: IBM looked at W3C Sofware license terms: so far looks good. Next
step is to put things in place within IBM.
? 2003-04-03: Editors to include normative schema language in
spec (conformance section?); schema to be
separate, in TR space.
PENDING 2003-04-03: Jonathan to respond to OWL with "no plans to
review, no resources, little knowledge of
why requested".
DONE [.3] 2003-04-03: Philippe to bring proposal to attention of M
Baker, P Prescod, R Costello, forward responses
to group
DONE [.4] 2003-04-03: Philippe to supply some example syntax for HTTP
binding proposal.
--------------------------------------------------------------------
4. Administrivia
a. May FTF [.1].
Registration [.2] closes on Apr 15th.
Dead line for hotel is May 1st.
b. Sept FTF (looking for West Coast host).
<unknown>
c. Usage Scenarios [.3]. Jonathan's still pretending to track this.
<unknown>
d. QA contact, test lead recruitment.
<unknown>
e. TF status
MEP TF:
David: good progress with Amy this week. Hope will be able to make
progress next week with Gudge. Amy will try to get Don to join.
FnP TF:
<glen sent regrets>
f. XML Schema 1.0 Second Edition review [.4]
Philippe: We may want to provide input on XML Schema 1.0 Second
Edition.
Gudge: Nothing in it will adversely affect WSDL.
------------------------------------------------------------------
5. New Issues. Merged issues list [.1].
- none
--------------------------------------------------------------------
6. Issue 2: SOAPAction has been deprecated, as of SOAP 1.2 [.1].
Jean-Jacques proposal at [.2].
Jacek's addendum at [.3].
proposal from Glen at [.4].
Topic: Property-based action syntax
Jean-Jacques: Update on SOAP 1.2. There is a URI for action.
+ the following fixes in a bug in Glen's proposal:
http://lists.w3.org/Archives/Public/www-ws-desc/2003Apr/0022.html
Tom: We shouldn't remove it since it is in WSDL 1.1 since new property
syntax is complex
Amy: Disagree. Shouldn't make HTTP a special case.
Tom: OK to do this both ways. Old way is "syntactic sugar".
Philippe: What do we gain by having a property syntax?
Amy: consistency
Jean-Jacques: It is possible to automatically transform the WSDL 1.1
syntax into the proposed PnF syntax
pro - consistency, con - ease of migration from WSDL 1.1
Jeffrey: good idea not to force change to existing, working syntax, but
also good to move forward
Tom: tools help authoring process but developers still need to read
WSDL to understand problems, e.g. in code generation from WSDL2Java.
PnF is more consistent syntax but requires knowledge of many URIs
Jacek: SOAPAction is a candidate for special treatment
Jean-Jacques: Web-Method is also likely to have to be represented as a
feature
Philippe: strawpoll:
Choice 1: Keep WSDL 1.1 syntax
Choice 2: use PnF
Choice 3: use WSDL 1.1 and PnF
X - William
A - Mike
A - Kevin
A - David
A - Philippe
A - SteveL
A - Dale
A - Prasad
A - SteveW
1 - Tom
1 - Arthur
1 or 2 - Umit
1 or 2 - JeffM
2 - Youenn
2 - Ingo
2 - Jacek
2 - Jerry
2 - Jean-Jacques
2 or 3 - Amy
3 - Lily
3 - Allen Brookes
3 - JeffreyS
3 - Martin
3 - Erick
A - 8
1 - 4
2 - 8
3 - 6
Philippe: no decision this week.
--------------------------------------------------------------------
7. BindingType proposal from Kevin [.1]. Updated proposal at [.2].
Topic: BindingType proposal
Kevin: Idea is to separate out binding detail into a separate construct
so it can be reused. syntax of binding and interface has stabilized so
let's reopen the discussion now. Here an updated proposal:
<bindingDetail name = "ncname" target = "binding|operation|message" ...>*
<!-- target attribute specifies where the details can be applied in
binding construct -->
<wsdl:documentation .... /> ?
<!-- binding details -->
</bindingDetail>
<binding name="ncname" interface = "qname" detail = "list of qnames" ...>*
<!-- the detail attribute may be a list of qnames, if conflicts
present , the first in the list takes preference -->
<wsdl:documentation .... /> ?
<operation name="ncname" detail = "list of qnames"? ...>*
<!-- if the detail attribute is not provided, it gets its value
from its parent. same below -->
<wsdl:documentation .... /> ?
<input name="ncname"? detail = "list of qnames"? ... >?
<wsdl:documentation .... /> ?
</input>
<output name="ncname"? detail = "list of qnames"? ...>?
<wsdl:documentation .... /> ?
</output>
<fault name="ncname"? detail = "list of qnames"? ...>*
<wsdl:documentation .... /> ?
</fault>
</operation>
</binding>
Gudge: Does this proposal incorporate the notion of layering,
e.g. defering detail. First define the use of SOAP and then reuse it
HTTP and SMTP for example?
Kevin: No.
Kevin: the bindingdetail is required and is the only place you can
specify binding. you must reference a bindingdetail from a binding.
JacekK: compare with Sanjiva's proposal. Kevin's is more reusable but
more complex than Sanjiva's, so let's revisit Sanjiva's:
http://lists.w3.org/Archives/Public/www-ws-desc/2002Jul/att-0117/01-bindings-2002-07-24.html
[Jacek presents Sanjiva's proposal]
Kevin: Is binding reusability worth the extra complexity
Tom: Reusability of binding details is less of a problem now that we can
"hoist" binding details - this reduces the amount of repetition
Philippe: so we have 3 "proposals": Kevin, Sanjiva, and Tom. Tom's one
doesn't preclude others though.
JacekK: pls clarify Tom's proposal
Tom: we now allow hoisting details, e.g. operation detail can be hoisted
to the interface, so there is no repetition within a binding, but also
no reuse across bindings
[Sanjiva joins the call]
Sanjiva: hoisting eliminates a lot of redundancy. providing defaults for
binding details would be a simplification. Another possible
simplification: if its a one-off binding, inline it inside the <port>
itself and avoid creating a named binding. That combined with defaulting
to doc/lit (er, I mean rpc/encoded ;-)) would go a long way to reducing
WSDL doc size
Philippe: close current discussion, put back on agenda next week
including Sanjiva's previous proposal and the idea of providing defaults
for bindings
Gudge: wasn't sanjiva's previous proposal tied in with serviceType?
Sanjiva: yes it was .. I just opened it up and its quite different. It
was focused on creating reusable bindings.
Umit: two issues: reusability and conciseness
Jeffrey: Attribute roll-up (old) and defaults (new) are related to conciseness?
ACTION: Sanjiva to rewrite his proposal on bindings
--------------------------------------------------------------------
8. Issue 6e [.1]
- Arthur's proposal on URL replacement [.2].
Plan to do the obvious cleanup suggested by Arthur, and fork the
rest into an R85 solution.
[pending R85]
--------------------------------------------------------------------
9. Issues 53-55 [.1]
- Philippe's proposal for HTTP binding [.2].
Philippe: Not enough time for adequate review. defer to next week.
Sanjiva: extend replacement to SOAP Action URI
[No takers to write a proposal to extend URLReplacement to SOAPAction
URI. the SOAPAction idea is tied into the binding defaults so sanjiva
will address this at his discretion]
[meeting adjourned. Don't forget to register for May f2f.]
Received on Thursday, 10 April 2003 14:01:46 UTC