- From: Jonathan Marsh <jmarsh@microsoft.com>
- Date: Wed, 25 Feb 2004 10:13:59 -0800
- To: <www-ws-desc@w3.org>
Minutes: Web Services Description WG
19 February 2004
Present:
Erik Ackerman Lexmark
David Booth W3C
Allen Brookes Rogue Wave Software
Paul Downey British Telecommunications
Youenn Fablet Canon
Martin Gudgin Microsoft
Hugo Haas W3C
Tom Jordahl Macromedia
Jacek Kopecky Systinet
Sandeep Kumar Cisco Systems
Kevin Canyang Liu SAP
Jonathan Marsh Chair (Microsoft)
Dale Moberg Cyclone Commerce
Jean-Jacques Moreau Canon
Bijan Parsia University of Maryland MIND Lab
Arthur Ryman IBM
Adi Sakala IONA Technologies
Igor Sedukhin Computer Associates
William Vambenepe Hewlett-Packard
Asir Vedamuthu webMethods
Sanjiva Weerawarana IBM
Umit Yalcinalp Oracle
Prasad Yendluri webMethods, Inc.
Regrets:
Roberto Chinnici Sun Microsystems
Glen Daniels Sonic Software
Amelia Lewis TIBCO
Ingo Melzer DaimlerChrysler
David Orchard BEA Systems
Jeffrey Schlimmer Microsoft
Jerry Thrasher Lexmark
--------------------------------------------------------------------
Agenda
1. Assign scribe. Lucky minute taker for this week is:
Igor Sedukhin (fallbacks: Jeffrey Schlimmer,
Dietmar Gaertner, Umit Yalcinalp, Jean-Jacques Moreau,
Sanjiva Weerawarana, Youenn Fablet, David Orchard)
Scribe: Jean-Jacques Moreau.
IRC: http://www.w3.org/2004/02/19-ws-desc-irc
--------------------------------------------------------------------
2. Approval of minutes:
- Jan 22nd telcon [.1]
- Jan 28-30 FTF [.2, .3, .4] and Summary [.5]
- Feb 5th telcon [.6]
- Feb 12th telcon [.7]
All minutes approved.
[.1] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0061.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0010.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0012.html
[.4] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0011.html
[.5] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0013.html
[.6]
http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/att-0035/040205-
ws-desc-irc.htm
[.7] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0076.html
--------------------------------------------------------------------
3. Review of Action items [.1].
PENDING 2003-09-18: Marsh to review the QA operational
guidelines.
PENDING 2004-01-08: Pauld to write up examples of schemas for the
Primer.
PENDING 2004-01-28: Philippe and JMarsh will look at the ipr for
test suite.
DUE FTF 2004-01-28: Sanjiva to consistify the @name attributes.
DUE FTF 2004-01-29: David Booth to suggest improvements to the
spec clarifying "WSDL processor".
PENDING [.2] 2004-01-30: DaveO to write up a proposal for augmenting
schema information to enable versioned data.
DONE [.3] 2004-01-30: DavidO to write request to schema group to
address the issue of schema not supporting
ignoring extended content.
DUE FTF 2004-01-30: Umit to write a proposal on @wsdlLocation
PENDING 2004-01-30: Jonathan to investigate typo in last f2f
meeting on _S_erviceType.
REASSIGNED TO HUGO 2004-01-30: Hugo to draft a note for the group around
safe operations.
DONE [.4] 2004-02-12: Philippe to check on teleconference facilities
for Tech Plenary f2f.
DONE [.5] 2004-02-12: Editors to update messageReference -> label in
part 2 as well.
DONE [.6] 2004-02-12: Issue list editor to make this a Part III issue.
Jonathan took up issues list editing
PENDING 2004-02-12: DaveO to produce a refined proposal for Asynch
HTTP binding addressing the concerns of folks
that object to leaving replyTo info out of WSDL.
DUE FTF 2004-02-12: Umit to update OperationName proposal to make
clear that this feature is always required.
DONE [.7] 2004-02-12: Jonathan to add links from the home page to the
edtodo and the media-types archive.
PENDING 2004-02-12: David Orchard to produces a specific example of
the kind of specification improvements he
envisions.
[.1] http://www.w3.org/2002/ws/desc/#actions
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0047.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0124.html
[.4] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0076.html
[.5] http://lists.w3.org/Archives/Member/w3c-ws-desc/2004Feb/0022.html
[.6] http://www.w3.org/2002/ws/desc/2/06/issues.html#x130
[.7] http://www.w3.org/2002/ws/desc/
--------------------------------------------------------------------
4. Administrivia
a. Upcoming FTFs
- March 4-5, Cannes-Mandelieu, France [.1]
Joint session with the TAG, XMLP?
No objection to have (informal or formal) meeting with XMLP at plenary
May happen during lunch
b. Web Architecture Document [.2, .3] review:
Volunteers so far: Jacek, Bijan, Jonathan
Next weeks call.
c. Charter renewal - everyone needs to be reappointed [.4]
AC reps need to reappoint WG participants
d. Handling Privacy [.5] review
Post FTF.
[.1] http://www.w3.org/2003/08/allgroupoverview.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2003Dec/0029.html
[.3] http://www.w3.org/TR/2003/WD-webarch-20031209/
[.4] http://lists.w3.org/Archives/Member/w3c-ws-desc/2004Feb/0033.html
[.5] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0119.html
------------------------------------------------------------------
5. Task Force Status.
a. Properties and Features (dormant)
b. Patterns (dormant)
c. Attributes (dormant)
d. Media type description (dormant)
XMLP worried pulled TF back into WSD WG
XMLP guys would still like to participate.
Please use the TF list for discussions on the proposal.
e. QA & Testing
- Response to comments on QA Spec Guidelines [.2]
- Implement QA Operational guidelines? [.3]
[.1] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0000.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2003Sep/0074.html
[.3] http://lists.w3.org/Archives/Member/w3c-ws-desc/2003Sep/0023.html
------------------------------------------------------------------
6. New Issues. Issues list [.1].
- 139: Non-deterministic schema (Gudge) [.2]
- 140: Version attribute proposal (Tom) [.3]
- 141: Should WSDL say anything about whitespace in SOAP:Body?
(Jacek) [.4]
- 142: Name of "message" component (Bijan) [.5]
- 143: Referencing other type systems (Bijan) [.5]
- 144: Why can't message reference simpleTypes? (Bijan) [.5]
- 145: How can you tell which type system is in use? (Bijan) [.5]
- 146: should WSDL be able to describe an operation with *anything*
in the message? (Jacek) [.6]
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0045.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0049.html
[.4] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0083.html
[.5] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0046.html
[.6] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0110.html
Note that I am not treating the following thread as an issue yet:
- Reuse faults by ref (DaveO) [.7]
I'm waiting for confirmation from DaveO that this is not obsoleted by
FTF decisions.
[.7] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0140.html
------------------------------------------------------------------
7. Issue 140: Version attribute [.1]
- Tom's initial proposal [.2] and follow-on proposal [.3]
(If this issue isn't ripe for decision, we'll move on fairly
quickly.)
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x136
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0049.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0069.html
Bijan: use Dublin-core
Sanjiva: no
Sanjiva: issues with inheritance, binding versioning, etc.
[* asir: Bijan, may I request you to give me a pointer to Dublin-Core
versioning mechanism.]
[jjm +1 to Sanjiva]
Tom: sympathetic to Sanjiva; however, like simple things; don't
want
to have complex mechanism
[Gudge: +1 from Gudge for a simple mechanism or no mechanism]
Paul: branching becomes important when more than one person
implementing
the same WSDL
Tom: Assumption: myWSDL, myService, indicate when things have
changed
in compatible or incompatible way
[Bijan: Dublin Core defines a metadata elemtn "relation". Relation has
a number of specializations inclusing isVersionOf,
isReplaceBy,
etc. See:
http://dublincore.org/documents/usageguide/qualifiers.shtml]
[*dbooth: wonders what Tom does in the case of a *mostly* compatible
change.]
[Bijan: There is a schema defined by DCMI:
http://dublincore.org/schemas/xmls/.]
Umit: Supported Tom's proposal at f2f. Need extra rules to support
multiple inheritance
Tom: no longer true in my latest proposal. Just use the leaf
interface version number, forget about parents
[Bijan: Some info on expression qualified dc:elements:
http://dublincore.org/documents/dc-xml-guidelines/]
Jacek: Problem with inheritance, e.g. B inherits from A, A changed
by owning company, then B, implemented by a different company,
is broken (as well as its binding)
[Bijan: I should note that what I do with dublin core is mostly in
RDF,
not in XML.]
Tom: No, by definition
Jonathan: back to mailing list
[pauld: thinks Dublin Core looks interesting, if Version can be
xs:string]
[Bijan: A discussion on using Dublin Core with XSLT:
http://www.biglist.com/lists/xsl-list/archives/200302/msg00251.html]
[Bijan: An article about using an RDF encodign of Dublin Core with
XSLT:
http://www-106.ibm.com/developerworks/xml/library/x-tipxrdf.html]
[Bijan: Hmm. I note a problem with using dublin core for 'version'...
it doesn't expres that document X *is version* (number) 1, or
2. It allows you to express that this document *is a version*
of some other document. How annoying.]
[Bijan: http://purl.org/dc/elements/1.1/identifier
[* JacekK asks bijan if he meant identifier for versioning and whether
dcterms:hasVersion vould be better (I don't know much about
dc, especially dcterms:hasVersion)]
[* bijan originally meant hasVersion and isVersionOf, but they are
*relational*. They say that I'm a version of X, not *which*
version I am. Identifier would allow associating a *Version
identifier* with a document.]
------------------------------------------------------------------
8. Issue 136: proposal to add in-optional-out [.1]
- Amy calls for a quick decision [.2]
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x136
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0042.html
Amy: +1
Jonathan: straw poll?
TomJ: mandatory to implement?
Jonathan: no
RESOLUTION: Accepted, no objection
------------------------------------------------------------------
9. Issue 96: Intermediaries [.1]
Jean-Jacques to re-present the concrete issue.
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x96
JJM: SOAP does support intermediaries. Intermediaries sit between
the initial sender and ultimate receiver. They can process and
add header blocks.
[* JacekK intermediaries can change the body.]
JJM: This processing can be described in a WSDL separate from the
ultimate receiver WSDL. But there is some interaction and
overlap.
Sanjiva: Could you come up with a concrete proposal? Seems to apply to
SOAP only; difficult to see exactly what this is about.
JJM: Yes, was actually planning on doing so this week; could do
next week. Supporting intermediaries has impact on client
and server side.
Umit: yes, and we care for both (in general).
Jonathan: back to email, and would be good if JJ could come up with
proposal
------------------------------------------------------------------
10. Issue 115: Improving on-the-wire conformance [.1]
Need a volunteer to develop a proposal.
Jacek did so (?) at [.2]
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x115
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0121.html
DavidB: remove definition of WSDL processor from WSDL language
Jonathan: is proposal from Jacek mutually exclusive [with issue 79]?
Jacek: My proposal does not affect conformance, just what a processor
should do in case of error.
Jonathan: can we merge them?
DavidB: Yes, will do
Umit: issue at Oracle with wire conformance. Cannot even talk about
wire conformance, for example if security is involved,
messages
on the wire will look different. What does a WSDL message
describe?
Sanjiva: it's abstract.
Jonathan: in the absence of extensions and features, what goes on the
wire is what is described in the binding.
Umit & Sanjiva: yes!
Jonathan: so no issue in that case?
Sanjiva: proposal for Part 3 ednote: "Actual binding format is not just
defined by the binding, but also by the extension elements in
use."
DavidB: Extensions may change the semantics of a WSDL document; for
that part of the WSDL document, gets delegated to that
extension.
Jonathan: so add something along the lines of Sanjiva's proposal? And
reclassify as Part 3?
Umit: Would like to see write-up until sees the issue
Sanjiva: make it editorial?
Umit: Need to check within Oracle
DavidB: May not be Part 3 only
Sanjiva: Part 1 doesn't talk about wire, etc.
Jonathan: need clear statement of our intent before closing the issue
Jacek: Should we say only required extensions may change semantics?
[Marsh: 1) Part 1 says extensibility changes the semantics; 2) Part 3
says Actual binding format is not just defined by the binding,
but also by the extension elements in use;]
DavidB: this is very valid, but issue 79
ACTION: Editor to add statement to Part 1 "extensibility changes
semantics".
ACTION: Editor to add statement to Part 3 "actual binding format
defined by extensions if present"
Confirm this proposed resolution at a later time.
------------------------------------------------------------------
11. Issue 143: Referencing other type systems (Bijan) [.1]
Text in 2.4.1 and 2.4.3 is inconsistent.
Option 1:
Clarify that extension type systems reuse the {message} component,
(optionally) add a type system identifier property.
Option 2:
Clarify that extension type systems add corresponding extension
Components, and {message} is XML Schema specific.
[.1] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0046.html
jmarsh summarises from agenda and asks bijan for a preference
Jonathan: need more flexibility?
JacekK: we are introducing extensibility for other type systems.
Problem
in bindings, e.g. SOAP binding. SOAP message could be
described
either by XML schema or SOAP data model schema. Propose to
suggest a message is a single XML element, defined using a
specific type system. Against extending messages as proposed
recently.
Jonathan: would issue 144 affect your decision on 143?
JacekK: simpleTypes can be treated as elements with simple content
Gudge: Not sure understands exactly the issue. Spec currently
doesn't require us to say where this elements came from
Bijan: it does say so
Sanjiva: this is a bug
Gudge: we indicate the element type could be populated by XML Schema,
but also by DTD or RelaxNG
Jonathan: some editorial work on section 2.2.1. Bijan would be satisfied
with this resolution to this issue.
[Gudge: Add text to row 3 of table 2-4 clarifying that the message
attribute ALWAYS refers to an element declaration in the
{element declarations} property on the definitions component
defined in 2.1. Clarify similar text in bulleted list in
Section 2.4.1]
Jacek: If allow adding other components, would need extra formalism
to mark these components as "data carrying", and binding is
expected to serialize such components.
[Bijan: Gudge: when would one populate a message component get
populated
by a non-element attribute? "The element declaration resolved
to
by the value of the message attribute information item if
present,
otherwise a similar construct in some type system as referred
to
by some other attribute information item if present, otherwise
empty."
[Gudge: If you used an attribute other than message.]
[Bijan: When would you do that?]
[Gudge: So mytypesystem:Construct='Java.Lang.String']
Jonathan: add note indicating in this case? cannot reuse existing
binding,
[Bijan: But that isn't an element declaration.]
[Gudge: Agreed. If you have an element declaration, you use
message='foo:bar']
[Bijan: So it couldn't populate a message component.]
JacekK: constrain WSDL to services that carry XML data
Jonathan: isn't this another issue?
[Bijan: So, *no* other such attribute information item will populate a
message component.]
JacekK: Motion to constrict ourselves to XML Elements on the grounds
of the alternative being too hairy
Jacek: restrict ourselves to simple elements
Jacek: came to the conclusion that any data model can be mapped to
XML
Sanjiva: Gudges' proposal seems to work; however some impact on style,
for
Example.
bijan: as one who will be using a radically different type system,
would like to understand what I will need to do to use this
different type system
JacekK: have to go; but no objection if close now
Gudge: Make life harder for binding if allow type system that don't
describe XML. Agreed earlier to support non-XML type systems;
but also to support non-XML bindings. The issue is at the
crossing: non-XML type systems with XML bindings.
Umit: had similar arguments about RPC. Was worried "element" would
no longer point to element type any longer.
Sanjiva: no, we're not doing that, AFAIU
Marsh: have we good statement if minutes?
Gudge: think so; could post by email
ACTION: Gudge: Send statement re. issue 143
Meeting adjourned.
Received on Wednesday, 25 February 2004 13:14:01 UTC