W3C

Web Services Description Meeting

11 Jan 2007

See also: IRC log

Attendees

Present
Charlton Baretto, Adobe Systems
Allen Brookes, Rogue Wave Software
Roberto Chinnici, Sun Microsystems
Jacek Kopecky, DERI Innsbruck at the Leopold-Franzens-Universität Innsbruck, Austria
Philippe Le Hegaret, W3C
Jonathan Marsh, Co-chair/WSO2
Monica Martin, Sun Microsystems
Jean-Jacques Moreau, Canon
Gilbert Pilz, BEA Systems
Tony Rogers, Co-chair/Computer Associates
Asir Vedamuthu, Microsoft
Regrets
Youenn Fablet, Canon
Tom Jordahl, Adobe Systems
Amelia Lewis, TIBCO
Arthur Ryman, IBM
Chair
Jonathan
Scribe
plh

Contents


 

http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/test-suite/Dashboard.html

http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/test-suite/results-messages/MessageTest-4G/log-MessageTest4G-canon-canon-results.html#message1

http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl20/wsdl20-adjuncts.html?content-type=text/html;%20charset=utf-8#_http_operation_multipart_encoding

"This format serializes the instance data in the HTTP message body, making it only suitable for HTTP requests using methods allowing message bodies."

http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/wsdl20/wsdl20-adjuncts.html?content-type=text/html;%20charset=utf-8#_http_operation_multipart_encoding

"This format serializes the instance data in the HTTP message body, making it only suitable for HTTP requests using methods allowing message bodies."

"In this serialization, for HTTP requests, the rules for constructing the HTTP request IRI defined in 6.7.1 Serialization of the instance data in parts of the HTTP request IRI apply if the {style} property of the Interface Operation bound has a value of "http://www.w3.org/@@@@/@@/wsdl/style/iri" as defined in 4.2 IRI Style."

http://dev.w3.org/cvsweb/~checkout~/2002/ws/desc/test-suite/results-messages/MessageTest-2G/log-MessageTest2G-canon-canon-results.html#message17

<Jonathan> http://lists.w3.org/Archives/Public/www-ws-desc/2007Jan/0058.html

<monica> will only be able to join today via irc - conflicts

<scribe> Scribe: plh

minutes from last week

Minutes approved

Action items

Review of Action items [.1].

[Interop]
?         2006-11-30: [interop] John Kaputin to create a test case 
                      with "required=false". 
?         2006-12-14: [interop] Jonathan to fix transferCodings - 
                      add control group

[WG]
?         2006-09-21: Jonathan to check periodically that SPARQL has 
                      added schemaLocation.
?         2006-12-14: plh to come up with a more detailed proposal for 
                      CR112 if possible
?         2007-01-04: Jonathan to make sure parameters are added to the 
                      URI in the HTTP binding even when no whttp:location 
                      appears.
?         2007-01-04: Paul to report back on which test cases in the 
                      WSDL test suite fail the basic patterns, with
                      suggestions on how to address the issues.
DONE [.3] 2007-01-04: Charlton to come up with a list of editorial 
                      comments and WSDL substantive issues.
?         2007-01-04: Jonathan to analyze CR117 further.

Current Editorial Action Items

Note: Editorial AIs associated with LC issues recorded at [.2].

[.1] http://www.w3.org/2002/ws/desc/#actions
[.2] http://www.w3.org/2002/ws/desc/5/cr-issues/actions_owner.html

<scribe> [DONE] 2007-01-04: Charlton to come up with a list of editorial comments and WSDL substantive issues.

Draft comments for WS-Policy

<Jonathan> http://lists.w3.org/Archives/Public/www-ws-desc/2007Jan/0071.html

Jonathan: let's skip the editorial issues
... let's look at number 11, policy intersection

Charlton: this is relevant to the client, not strictly relevant to WSDL. Don't see anything to select the intersection mode selection.
... if one entity understands one mode and an other is being required by the other entity, the intersection won't occur.

Jonathan: so, if I would attach a policy to WSDL, should we have a mode as well?

Charlton: if you want someone to always apply lax, there is no mechanism for that for the moment.
... there should be a mechanism for expressing this.

Philippe: this is up to the client to use one or the other intersection mode

Asir: it's at the discretion of the requester. I don't see this as a WSDL issue

Charlton: not trying to express that anyone other than the client can pick the mode, but the service might want to indicate which one to pick.

Jonathan: be isn't ignorable the mode?

Philippe: if you don't want the client to ignore your assertions, then don't use ignorable.

Jonathan: we would need a more developed use case to see the interop problem, but this seems a policy issue

Monica: some parts of the discussion will be handled in other documents in WS-Policy. Maybe that will help Charlton on this issue.

Charlton: I'm happy not including this in the WSDL comments.
... Number 12 is already filed as an issue in WS-Policy but we're expressing interest in it.
... the issue is about how to map parameters in policy attachments to WSDL 2.0.

Asir: is this a Last Call issue or an issue about the WSDL 1.1 Element identifiers?

Charlton: this is a last call issue?

s/\?//

scribe: it's a fragid issue.

Jonathan: I suggest we include it in the mail, as a side remark.
... I'll file the issues on behalf of the WSDL WG.
... all editorials. not 11.
... I'll leave the observations in this WG.

<scribe> ACTION: Jonathan to send WS-Policy comments to the WS-Policy WG [recorded in http://www.w3.org/2007/01/11-ws-desc-minutes.html#action01]

Interchange format draft note

<scribe> Postponed

MTOM Description

<Jonathan> http://www.w3.org/Submission/2006/SUBM-WS-MTOMPolicy-20061101/

Jonathan: W3C ack'd a submission on the MTOM policy assertion

Philippe: will be sent to XMLP. I *think* they want to work on this on the REC track.
... the XMLP WG will need a new charter if they do so.

Jonathan: following-up on this, I thought Canon was going to investigate the use of the MTOM policy assertion. Any news?

Jean-Jacques: no news yet.

Jonathan: the interesting point is if it is difficult to implement a WS-Policy that can only support one policy assertion.

Jean-Jacques: if we parse it, and check if the infoset only contains a subset, would that work?

Jonathan: you also want to support the PolicyReference element, which is the practice
... you also want to reject policy you don't understand.

Jean-Jacques: a bit more complex than XPath...

Jonathan: I also don't understand the resources constraint you're facing
... could one use XSLT to transform the policy into a direct WSDL extension?
... if that's true, the information content is equivalent
... Paul's proposal was to annotate the policies. Asir's proposal was that we don't need to annotate them at all.

<scribe> ACTION: Jean-Jacques to provide more analysis on how difficult it would be deal with a Policy that only contains an MTOM policy assertion [recorded in http://www.w3.org/2007/01/11-ws-desc-minutes.html#action02]

CR122

http://www.w3.org/2002/ws/desc/5/cr-issues/#CR122

John: 6.7.2.2.2 talks about ignoreUncited. If the HTTP PUT or POST then the remainder gets serialized in the message body anyway, not repeated in the URI. So the use of the property isn't relevant here.

Jonathan: should we clarify that data could be thrown away?
... you're not only preventing the data from appearing in the URI, but you can also prevent it from appearing in the HTTP Body
... Youenn got the correct behavior
... I do believe it would be possible to write some WSDL that wouldn't interop because of implementation limitations.
... yes, it is the expected behavior. The user might trip on this. Closed with no action.

Resolution: close with no action

CR130

http://www.w3.org/2002/ws/desc/5/cr-issues/#CR130

<monica> nick /monica

John: I had to make some decision on how to associate the tokens I find with the string
... should the spec be more specific?

<jkaputin> whttp:location="{{{town}}}"

<jkaputin> {{,{town},}}

<jkaputin> {{,{,town,}},}

John: two ways to interpret the example.

<Roberto> I'd say it's '{' + value of town + '}'

+1

John: with a strategy that looks for matching par curly braces first, the example will be ok.

Jonathan: the proposal is to match the pair of curly braces first.

Tony: I can't see a grammar where the second result would be correct

Jonathan: is there a common algorithm we can point to?

John: don't know any

Philippe: don't try using regular expression to handle http:location, you need a parser and a state table

{{town}}

<scribe> ACTION: Philippe to propose a grammar for http:location [recorded in http://www.w3.org/2007/01/11-ws-desc-minutes.html#action03]

CR131

http://www.w3.org/2002/ws/desc/5/cr-issues/#CR131

John: should we deal with infault elements as well?

Jonathan: the binding only works for the 3 MEPs
... you can say that it applies to the infault in your extension as well.

Resolution: the HTTP 1.1 Binding aplies only to the 3 MEPs we have in the spec. Closed with no action

CR 133

http://www.w3.org/2002/ws/desc/5/cr-issues/#CR133

Jonathan: For the SOAP response MEP the immediate destination is "the value of the WSDL {address} property, modified by the {http location} property following the rules described in section 6.7.2 Serialization as application/x-www-form-urlencoded."
... is it intentional that we are ignoring http:location?

"the SOAP request-response MEP the immediate destination is "the value of the WSDL {address} property of the Endpoint component."" . is it intentional that we are ignoring http:location?

John: so instance data would be serialized in the IRI

Jonathan: yes, but the full XML would still be part of the SOAP Body

John: might be useful in gateway/middleware scenarios
... so you don't have to parse the message body but still want to do dispatch

Jonathan: didn't see any reason to have it the way it is.

Proposal: the SOAP request-response MEP immediate destination is the value of the

WSDL {address} property, modified by the {http location} property following the

rules described in section 6.7.2 Serialization as application/x-www-form-urlencoded. The entire instance data is serialized in the SOAP Body.

Resolution: proposal adopted.

CR 140

John: What happens if the MEP allows multiple input messages or infaults as well as input messages?
... there is an issue on how to associate the element name with the input data.

Jonathan: we only define the property for the 3 MEPs.

Resolution: closed.

CR 141

http://www.w3.org/2002/ws/desc/5/cr-issues/#CR141

Jonathan: this is a duplicate of 130 since Philippe will come up with a grammar.

Resolution: will be resolved with the resolution of 130. Closed (duplicate).

CR119

wsdl: binding/wsdl:operation/@whttp:queryParameterSeparator is missing in

the HTTP binding syntax summary while present in section 6.4.4.

Resolution: agreed. needs to be fixed.

CR 120

Jonathan: It does make sense to add #none
... the SOAP Response MEP can only be used with the IRI style or with an input message of #none.
... sounds reasonable.

Resolution: agreed. Needs to be fixed.

CR 121

Support for deprecation.

Resolution: could be done through extensibility or in V.Next. Closed.

Summary of Action Items

[NEW] ACTION: Jean-Jacques to provide more analysis on how difficult it would be deal with a Policy that only contains an MTOM policy assertion [recorded in http://www.w3.org/2007/01/11-ws-desc-minutes.html#action02]
[NEW] ACTION: Jonathan to send WS-Policy comments to the WS-Policy WG [recorded in http://www.w3.org/2007/01/11-ws-desc-minutes.html#action01]
[NEW] ACTION: Philippe to propose a grammar for http:location [recorded in http://www.w3.org/2007/01/11-ws-desc-minutes.html#action03]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2007/01/11 17:40:29 $