- From: Jonathan Marsh <jmarsh@microsoft.com>
- Date: Fri, 13 Feb 2004 12:35:40 -0800
- To: "WS Description List" <www-ws-desc@w3.org>
WSDL WG Telcon
12 February 2004
Attendence:
David Booth W3C
Allen Brookes Rogue Wave Software
Roberto Chinnici Sun Microsystems
Glen Daniels Sonic Software
Paul Downey British Telecommunications
Youenn Fablet Canon
Tom Jordahl Macromedia
Jacek Kopecky Systinet
Philippe Le Hégaret W3C
Amelia Lewis TIBCO
Kevin Canyang Liu SAP
Jonathan Marsh Chair (Microsoft)
Jeff Mischkinsky Oracle
Dale Moberg Cyclone Commerce
David Orchard BEA Systems
Bijan Parsia University of Maryland MIND Lab
Arthur Ryman IBM
William Vambenepe Hewlett-Packard
Asir Vedamathu webMethods, Inc.
Sanjiva Weerawarana IBM
Umit Yalcinalp Oracle
Prasad Yendluri webMethods, Inc.
Regrets:
Ingo Melzer DaimlerChrysler
Erik Ackerman Lexmark
Ingo Melzer DaimlerChrysler
Jean-Jacques Moreau Canon
Jeffrey Schlimmer Microsoft
--------------------------------------------------------------------
Agenda
1. Assign scribe. Lucky minute taker for this week is:
Igor Sedukhin (fallbacks: Jeffrey Schlimmer, Prasad Yendluri,
Dietmar Gaertner, Umit Yalcinalp, Jean-Jacques Moreau,
Sanjiva Weerawarana, Youenn Fablet, Youenn Fablet, David Orchard)
Scribe(s): Prasad and Sanjeva
<dbooth2> Philippe: I'm trying to have the WSDL charter this week, and include the SOAP 1.1 binding with http put and delete (but not required).
<dbooth2> ... Also, I'll be stepping down as team contact and Hugo Haas will replace me. I'll still be at the Tech Plenary though.
<dbooth2> JMarsh: Thanks for all the work you've done.
<dbooth2> Philippe: The change will become effective when the charter is effective.
<dbooth2> dbooth: I'll be continuing as alternate team contact.
--------------------------------------------------------------------
2. Approval of minutes:
- Jan 22nd telcon ?
- Jan 28-30 FTF [.2, .3, .4] and Summary [.5]
- Feb 5th telcon [.6]
[.1] ?
[.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
Umit: I did not have a chance to review the F2F minutes.
JM: What about last week's telcon minutes.
Umit: Did not review those as well
JM: Ok. We will carry minutes approval to next week.
--------------------------------------------------------------------
3. Review of Action items [.1].
PENDING 2003-09-18: Marsh to review the QA operational
guidelines.
DONE [.2] 2003-11-04: Glen to write up rationale for removing headers
(and?) proposal for a generic header-adding
property/feature.
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.
PENDING 2004-01-28: Sanjiva to consistify the @name attributes.
PENDING 2004-01-29: David Booth to suggest improvements to the
spec clarifying "WSDL processor".
PENDING 2004-01-30: DaveO to write up a proposal for augmenting
schema information to enable versioned data.
PENDING 2004-01-30: DavidO to write request to schema group to
address the issue of schema not supporting
ignoring extended content.
DONE [.3] 2004-01-30: Tom to write proposal on version attribute
(modeled after schema's).
PENDING 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.
DONE [.4] 2004-01-30: Issues list editor to retitle Issues 113 and
91.
PENDING 2004-01-30: Philippe to draft a note for the group around
safe operations.
[.1] http://www.w3.org/2002/ws/desc/#actions
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0053.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0049.html
[.4] http://www.w3.org/2002/ws/desc/2/06/issues.html
--------------------------------------------------------------------
4. Administrivia
a. Upcoming FTFs
- March 4-5, Cannes-Mandelieu, France [.1]
Joint session with the TAG, XMLP?
JM: Any questions? I will add an agenda item for call the wk before F2F
reg review of the Arch doc and any comments/issues to be brought up
with TAG
Amy: Telcon facility available (at the tech plenary?)?
Philippe: I believe so. I need to check
JM: Anyone wants to dial in, post to admin list that you want to register for that.
We will schedule a call based on how many people and at what times.
ACTION: Philippe to check on teleconference facilities for Tech Plenary f2f
[See later in these minutes for results of this action item.]
b. Web Architecture Document [.2, .3] review:
Volunteers so far: Jacek, Bijan, Jonathan
[.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/
------------------------------------------------------------------
5. Task Force Status.
a. Properties and Features (dormant)
b. Patterns (dormant)
c. Attributes (dormant)
d. Media type description (dormant)
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
Skipped.
------------------------------------------------------------------
6. New Issues. Issues list [.1].
- Issues 137, 138 already added (see below)
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0137.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0022.html
[.4] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0034.html
Note that I am not treating the following thread as an issue yet:
- Reuse faults by ref (Daveo) [.5]
I'm waiting for confirmation from DaveO that this is not obsoleted by
FTF decisions.
[.5] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0140.html
JM: I did not add a couple of new ones but, 137 and 138 are already added to the issues list. Skipping down to issues list.
------------------------------------------------------------------
7. Ratifying renaming decisions from FTF.
FTF produced proposals with general acclaim to rename the following
constructs:
- Issue 118: s/@message/@element/ [.2]
- Issue 119: s/@messageReference/@label/ [.3]
- Issue 116: s/A,B/IN,OUT/ (already closed?) [.4]
[.1] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0011.html
[.2] http://www.w3.org/2002/ws/desc/2/06/issues.html#x118
[.3] http://www.w3.org/2002/ws/desc/2/06/issues.html#x119
[.4] http://www.w3.org/2002/ws/desc/2/06/issues.html#x116
Topic: Issue 118
Rename message @ to element
JM: Bijan as part of his new issue suggested we need to change the name
of the message component. We can keep component and @ name together.
Bijan: Message component in one section says it can take values from arbitrary
type system and in another part says it can only take element decl.
JM: Sounds like a bug. It should say element declarations from any type system.
JM: Lets take renaming message @ to element
Tom: I am in favor of changing the att name to element.
Arhur: What is the type? Is this going to be a QName? I prefer message as it
conveys more semantic information and we should just say its a QName.
Tom: Its already a QName
Umit: I agree with Tom and we should change it to element
Glen: Message conveys wrong semantics as it is part of a message in SOAP world becomes body.
DaveO: Element does not get us anywhere better
Glen: Element says we are referring to a schema element
Tom: Fits with WSDL 1.1 type vs element issue. Says we picked element in WSDL 2.0
JM: Anyone object to renaming? Arthur?
Arthur: Not a big deal. What is in a name?
DaveO: I won't object but, I express my reservations.
JM: We discussed this a lot at f2f. Last call to objections.
No objections. Issue resolved.
RESOLUTION: Issue 118 closed, accept proposal to rename attribute.
Philippe: Telcon update. We do not have telcon at the tech plenary except on Wed.
For any WG.
Amy: Who else needs telcon facility?
JM: All that want a telcon facility send a note to admin list
Topic: Issue 119
Rename @ messageReference to label
Bijan: I was mistaken. It was supposed to a reference to a placeholder message
Glen: It is an abstract message ref holder
Bijan: Just pointing out my original rationale was wrong
JM: Any concerns w/ renaming @ messageReference to label ?
RESOLUTION: issue 119 closed, renaming of attribute and associated editorial cleanup approved.
Topic Issue 116
Renaming message labels in the patters from A and B to IN and OUT
JM: From issues list and f2f minutes it seems we resolved we would rename as above
Amy: I already made these changes, as I was under the impression this was resolved
as editorial at the f2f.
Roberto: I like A and B better, A B and C imply the message order.
I don't remember the discussion at f2f
Tom: For patterns beyond simple in and out patterns I expect the in and out to have numbers
at the end of them if there are multiple ins and outs
Roberto: What about the order between Ins and Outs?
Glen: Implied ordering with As and Bs does not scale well
DBooth: If we reopen the issue, we could name them In-1 Out-2 etc.
Tom: Only if we re-open?
Dbooth: Yes. We are not reopening we can leave the current INs and OUTs
JM: Can we leave it closed and move on?
Amy: In part II there is a component called messageRef. Should that also be changed to label?
It provides further consistency. Any objections to doing that?
No objections. Editorial change.
ACTION: editors to update messageReference -> label in part 2 as well
Issue 116 closed
------------------------------------------------------------------
8. Issue 137: Properties should not be limited to simple types [.1, .2]
- Proposed resolution: Change the xs:anySimpleType in section 2.7.2.3
to xs:anyType, and appropriately rewording the text in table 2-7.
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x137
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0034.html
Topic: Issue 137
Properties should not be limited to simple types
JM: Glen proposed resolution.
Proposed resolution: Change the xs:anySimpleType in section 2.7.2.3
to xs:anyType, and appropriately rewording the text in table 2-7.
JM: Any concerns w/ accepting the proposal
RESOLUTION: Accept the proposal and close issue 137.
------------------------------------------------------------------
9. Issue 130: Need async request/response HTTP binding [.1]
- Two formulations at [.2, .3]
- STATUS: Need to gauge interest, action someone to refine the
pattern.
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x130
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0192.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0195.html
Topic: Issue 130
JM: This is actually part III issue not part I. We don't have really deal with it
prior to f2f. Five minutes
JM: There is some discussion on the list
DaveO: We had some discussion. We need to have proposal on the table that people can live with
JM: Still needs a lot of discussion. Anyone interested in championing a proposal along with BEA
Prasad: I am interested
Jacek: It is a bit complex to put in WSDL
DaveO: We need a more concrete proposal
A number of usecases where people want to layer asynchrony on top of HTTP
We need a standardized binding for a request that comes on one HTTP request
and a call back that happens later on, via another HTTP request.
Paul: This is of interest to us.
DaveO: We are not proposing that we standardize on WS-Addressing ReplyTo or
any other spec that defines where the call-back address the SOAP response goes to.
It will not be defined in the WSDL binding.
DBooth: DaveO how the Provider Agent knows the address of the Requester Agent
Paul: It could be dynamic and conveyed in the messages exchanged.
DBooth: So the reply address will be application defined and will not be in the WSDL syntax
Umit: So, are we going to define something that is not really complete? Then we will be forced
to the particular addressing scheme that the application wants.
DaveO: Don't we have the same issue with Out pattern?
Umit: Yes but, if someone is going to rely on the replyTo parameter then we need to define it in WSDL
and use it. Otherwise the pattern will not be usable.
DaveO: I agree that this is piece of info that is dynamic that is needed.
We already do this with Content-Location header in HTTP
This is a layering problem. We provide the infrastructure, people can layer on right dynamic behavior they want
Umit: But we need to a well defined property like "reply-destination" that we can rely on,
that is defined at the WSDL level
DaveO: We need a standard header but, WSDL is not the place for this.
Umit: Why not?
DaveO: If we do this, it opens up the bucket into whole different areas.
Jacek: This is the same incompleteness situation with operation dispatch
Glen: We need to identify in WSDL the fact that this information is needed
to use the binding, even though WSDL does not supply it. People can use their
own mechanism.
JM: We need to do couple of things. Change this from part I issue to part III issue
We need to get a more refined proposal on the table, after the F2F
ACTION: Issue list editor to make this a Part III issue
ACTION: DaveO to produce a refined proposal for Asynch HTTP binding addressing the
concerns of folks that object to leaving replyTo info out of WSDL
------------------------------------------------------------------
10. Issue 138: Second level xs:import [.1, .2]
- Proposed resolution [.3]
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0137.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Feb/0022.html
Topic: Issue 138
second level xs:import
JM: Proposal was to add some text clarifying this in the spec, that got
degenerated into argument over what is a normative text and what is not.
Umit: I agree with Yaron that it would not hurt to add a bit more text
to the spec to clarify this.
Amy: We cannot govern the semantics of how Schema does import. We can't redefine that
in WSDL spec. We cannot change the semantics of schema import.
Umit: That is not what is being requested. A bit more clarification text.
Amy: If we are adding non-normative text, I am in favor of explaining schema import
We got clarify that we can not change schema and any explanation we give is non-normative
Prasad: People of this WG have tripped over this issue a few times and I agree this deserves
clarification. I see Amy's point on non-normative text. We should consider putting this in
Primer.
JM: This is really a clarification issue
DaveO: Can we wait for week prior to closing this?
JM: Original proposal was to allow to refer to schema constructs that are buried
deep in the schema import chain
Tom: That is not going to fly. We decided to punt that to Schema.
JM: Yes. Moreover, the WSDL import is designed to be consistent with Schema import.
DaveO: This would be good topic to include in the primer
Umit: Putting in the primer or not, it does not change the semantics
JM: If we want to explain how this all works perhaps primer is a better place
JM: Dbooth, Do you want to put this in this primer.
DBooth: If a small non-normative text in the spec would do it, that is fine
Amy: If we put this in the spec, it needs to be clearly marked as informative
JM: we could put it as a NOTE
DBooth: Are Notes non-normative? I am not sure. Need to check.
It is editorial issue, how to mark this non-normative
JM: Ok lets accept the proposal to add explanatory text non-normatively to the core spec, with marked
as an editorial item to figure out where to put it in the spec, and how to mark it non-normative;
and close this issue. Any objections?
Arthur: Question, is this WSDL import or schema import issue?
JM: Specifically about referring from WSDL to schema components that are imported
Arthur: Imported via another WSDL ? WSDL permits two kinds of imports.
It is no clear what the chain of imports are.
JM: The schema element in types section imports another schema, from WSDL I can only
refer to schema components that are defined in the in-line schema unless I add a separate
import for the second schema at the WSDL level.
Amy: Essentially we need to obey the semantics of schema import
Arthur: If WSDL import is the source of confusion, we should make it more precise
Amy: Do we have an issue to clarify visibility of schema elements of imported WSDLs?
JM: I thought we addressed that at F2F and accepted a proposal.
Amy: Ok
JM: Ok 138 is successfully disposed (closed).
RESOLUTION: Close issue 138 by adding explanatory text to Part 1 explaining the behavior and the reason for it (suggestion in the proposed resolution mail).
------------------------------------------------------------------
11. Issue 120: Operation Name feature proposal [.1, .2]
Mark Baker had some comments [.3, .4].
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x120
[.2] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0082.html
[.3] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0173.html
[.4] http://lists.w3.org/Archives/Public/www-ws-desc/2004Jan/0175.html
New Topic:
Issue 120: Operation Name visibility feature proposal
Umit: I don't understand Mark Baker's issue is with the proposal.
Umit: Explains her proposal
Glen: It seems Mark Baker wants style instead of just features and properties
Umit: The proposal says if you are using RPC style you are automatically compliant
with the feature, because of the way RPC style is defined. RPC style is not a requirement
Glen: You could say style implements the feature
Umit: The problem we are trying to resolve does not occur with RPC style
Glen: MB says you can also introduce other styles like inherited, .... I don't know exactly what he wants :)
Umit: The intent of this whole thing is solve the uniqueness of an operation on the wire problem
Glen: We need more time to understand what MB is asking for before we can resolve this issue
Umit: People have been complaining we don't have good examples of features and properties.
This also defines two different SOAP modules.
Sanjiva: Why does this proposal have a module definition?
Glen: We need to deal with this in the doc/lit case
Sanjiva: There is no doc/lit concept anymore
...
clarifying discussion on features, properties and modules how they are used
....
Jeff: Rather than doing an exegesis on MB's proposal, we ought to say we don't understand MB's issues
and we can not make any changes until we understand.
Jacek: I like the proposal but, in the module you identify the operation by the component designator URI.
I think a QName would be better. It is a minor tweak.
Glen: It was originally QName but we changed. Can't recall why.
Umit: Operation names are not really QNames. This is perhaps an issue.
There are warnings in the current spec not to treat operation QNames names as unique.
Hence we needed to change that a URI.
Jacek: I withdraw my amendment
Dbooth: I am just wondering to what extent we want app developers from hanging themselves
Umit: This is not a question of preventing people from hanging themselves but, there is a real need to
convey the operation name in the message.
JM: Prasad needs to drop off in few minutes. We need a new scribe.
Sanjiva volunteers.
Prasad: Thanks Sanjiva
Sanjiva: There was requirement before that all the names in an interface should be unique. If they are
unique I don't see a need for this.
Umit: For tool vendors it is difficult to relate a message coming on the wire to an operation.
This feature makes it possible.
...more discussion on if we can live w/o this feature ....
Glen/Jacek: What if you are trying to implement a WSDL that someone else defined
and the signatures not unique?
<END of Prasad's Scribing>
<Sanjiva takes over>
<JacekK> Jacek supports the proposal by Umit and Glen
dbooth: proposes that WSDL generators have a rule saying they will
generate unique messages
glen: its not necessarily wsdl's problem ..
<JacekK> but if I take a WSDL (an interface) that doesn't make
messages unique on the wire and I'm implementing it, I can
use the header, but don't I have to change the interface to
specify it uses the feature anyhow? The basic idea was that
I wouldn't have to change the interface.
sanjiva (while he was scribe): there are other solutions to doing this
without WSDL having to address this; it may involve specs that
are not part of WSDL scope
JM: asks Glen whether its possible to define these features,
properties and modules as a separate doc
Glen: yes its possible
JM: Notes that we don't define any other specific modules or soap
headers right now.
Glen: explains further how a SOAP engine would do this via a module
Umit: the definition of the feature and its implementation can be
separated
Umit: would like to put soap module and feature definition in core spec
Sanjiva: proposes a new document to contain specific features such as
the operation name
Glen: this will be an always required feature
Sanjiva: That changes this discussion completel
Umit: No that's not required .. only if u put it in the WSDL
Glen: No it is always required
Umit: Will reply to Mark Baker's concerns to try to clarify his position
JM: Let's try to find a way to move forward
Glen: syntax of this feature doesn't necessarily have to be in every
WSDL, but this feature is always there conceptually
ACTION: Umit to update proposal to make clear that this always required
Sanjiva: Why did you not propose this to be a part of the WSDL namespace
if it must always be there?
Paul: Is this feature binding specific?
Glen/Umit: No but the soap:module is of course binding spceific
Jack: doesn't see why every WSDL must implement this feature?
Tom: If u support the RPC style then you automatically satisfy this
feature
JM: Proposes we think about this and let Umit fulfill her AI
------------------------------------------------------------------
12. Issue 135: WSDL Specification readability [.1]
"The WSDL specs contains a lot of formulaic text, making them
harder to read than they could be. A lot of the infoset related
data could be easily moved into an appendix. This would make the
bulk of the spec just as informative and a lot easier to read.
Could we move some of the formulaic text to an appendix?"
[.1] http://www.w3.org/2002/ws/desc/2/06/issues.html#x135
Sanjiva complains ..
JM: May be improvements that can be done to reduce formalaic stuff
DaveO: accurate representation (move infoset gorp to an appendix)
DaveO: can we re-factor to improve it to icrease the boilerplate/noise
ratio
DaveO: will come up with concrete examples of what can be factored
Sanjiva: says we did this discussion earlier and decided on the model
we have now
Glen: compromise: include some example
Sanjiva: +1 to adding examples; will you provide them please?
Glen: agrees!!!
DaveO: asks Sanjiva to summarize some of the variations of the spec
we've gone thru
Sanjiva: suggests looking at WD1 vs. WD2
JM: Asks DaveO to propose specific changes and suggests we
shouldn't embark on a re-factoring at this time.
Umit asks where the edtodo is:
* scribe umit its .../edtodo.html
http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl20/edtodo.html
ACTION: Jonathan to add links from the home page to the edtodo and the media-types archive.
Meeting adjourned.
Received on Friday, 13 February 2004 15:35:45 UTC