Web Services Resource Access Working Group Teleconference

15 Dec 2009


See also: IRC log


Alessio Soldano, Red Hat
Ashok Malhotra, Oracle Corp.
Asir Vedamuthu, Microsoft Corp.
Bob Freund, Hitachi, Ltd.
David Snelling, Fujitsu, Ltd.
Doug Davis, IBM
Fred Maciel, Hitachi, Ltd.
Gilbert Pilz, Oracle Corp.
Katy Warr, IBM
Li Li, Avaya Communications
Martin Chapman, Oracle Corp.
Ram Jeyaraman, Microsoft Corp.
Sreedhara Narayanaswamy, CA
Tom Rutt, Fujitsu, Ltd.
Vikas Varma, Software AG
Yves Lafon, W3C/ERCIM
Bob Natale, MITRE Corp.
Jeff Mischkinsky, Oracle Corp.
Mark Little, Red Hat
Orit Levin, Microsoft Corp.
Paul Fremantle, WSO2
Paul Nolan, IBM
Prasad Yendluri, Software AG
Wu Chou, Avaya Communications
Bob Freund, Hitachi, Ltd.
Katy Warr


<trackbot> Date: 15 December 2009

<Bob> scribenick: Katy

Appoval of agenda

dug: Pls can we talk about 8201 if possible?

Agenda agreed

Approval of minutes 8th Dec

RESOLUTION: No objects, minutes of 2009-12-08 approved

Publication of 17/11 snapshots

RESOLUTION: Approve publication of 17 Nov snapshots as heartbeat

<Bob> http://www.w3.org/2002/ws/ra/tracker/actions/open

Discuss action items

Bob: No meetings after this one until Jan 5th

<dug> I don't recall Gil getting Bob's permission to go to Australia

<dug> clearly, he was out of order

Bob: Presume 8176 will stay until next meeting.

Gil: 8284 Text in WSDL 1.1 not consistent with schema.

Bob: One of problems was BP restricted to HTTP and SOAP 1.1

Gil: Need carefully crafted text.


Asir: we have a bunch of ws-ra spec wsdls already and we can already build BP compliant.

Gil: We are talking about references to WSDL 1.1 in our specs when WSDL 1.1 is broken

Asir: but why does this matter as our WSDLs are ok

<asir> ... also our operation descriptions are okay

<gpilz> here's what I'm talking about: 4.7.12 Describing headerfault Elements There is inconsistency between WSDL specification text and the WSDL schema regarding soapbind:headerfaults. R2719 A wsdl:binding in a DESCRIPTION MAY contain no soapbind:headerfault elements if there are no known header faults. The WSDL 1.1 schema makes the specification of soapbind:headerfault element mandatory on wsdl:input and wsdl:output

Asir: Our WSDLs are BP compliant so why does it matter

Gil: We should reference a version of WSDL 1.1 that is BP compliant - where the inconsistencies have been addressed

<asir> WS-RA specs don't define any headers

<dug> so a BP compliant WSDL isn't WSDL 1.1 compliant?

Bob: How about we keep the normative ref to WSDL (same as BP) but mention that we would expect implementations to be conformant to the requirements of BP?

Asir: but wouldn't we need to list the requirements

Bob: It would be up to the implementer to choose
... depending on what they were doing

Asir: We would like to consider this

<scribe> ACTION: Gil to produce some specifics for 8284 that reflect the relevant aspects for ws-ra [recorded in http://www.w3.org/2009/12/15-ws-ra-minutes.html#action01]

<trackbot> Sorry, couldn't find user - Gil

<Bob> ACTION: Gilbert to produce some specifics for 8284 that reflect the relevant aspects for ws-ra [recorded in http://www.w3.org/2009/12/15-ws-ra-minutes.html#action02]

Issue 6463

Katy: Described 2nd proposal

Asir: We like the proposal but have some further changes marked up

<trackbot> Could not create new action (failed to parse response from server) - please contact sysreq with the details of what happened.

<trackbot> Could not create new action (unparseable data in server response: local variable 'd' referenced before assignment) - please contact sysreq with the details of what happened.

Asir: we can walk through now

<dug> I'd like more time to review Asir's proposed edits

<dug> http://lists.w3.org/Archives/Public/public-ws-resource-access/2009Dec/0067.html

<li> anybody having problem with http://lists.w3.org?

<dug> yes I'm having 'issues' too

<Yves> no problems should be with www.w3.org currently

Asir: Would like to call out assumption for 8.1

<dug> email site is ok- main w3 site is slow/down

<dug> e.g. bugzilla is down for me

Asir: and describe other details of the markup

Dug: Please could we have a marked up copy

Ashok: I made some comments on the syntax - in particular the layers of wrapping needed in the Metadata
... could be a new issue
... but this is related to how the policy is attached to the endpoint so I am wondering what to do with these concerns
... Issues are: 1) schema url is duplicated
... 2) We have 3 layers of wrapping

(08) Identifier='http://www.w3.org/2009/09/ws-mex'>

(09) <wsdl:definitions name='StockQuoteMetadataExchangeMetadata'

(10) targetNamespace='http://www.w3.org/2009/09/ws-mex'

(11) xmlns:wsdl='http://schemas.xmlsoap.org/wsdl/'

Asir: The above is just that the identifier is a hint for what is in the WSDL - the identifier is optional
... so is not required

(04) <wsa:Metadata>

(05) <mex:Metadata xmlns:mex='http://www.w3.org/2009/09/ws-mex'>

(06) <mex:MetadataSection

Ashok: Why not take out mex:Metadata?

Asir: Because the mex:metadata may be embeded in transfer response, epr or mex
... transfer requires one wrapper which is first wrapper in resource representation
... the consistent wrapper is useful for consistency

Dug: The mex arwpper in transfer is only useful in transfer when all metadata is wrapped in one blob
... if you are moving mex metadata documents it provide consistency
... but other metadata - wsdl, schema - does not have wrapper so ws-t, http do not get consistency through mex metadata
... I think common usecase would be 'give me your wsdl' where the mex:metadata is not require

Ashok: Agree
... We have not reached agreement on the points that I raised
... Issue is: The syntax of attaching mex matadata should be simplified

Bob: Would anyone object to opening an issue?

Katy: I would like to discuss Ashok's issue now so that we can get it in the open

Bob: As this is a substantive issue, my recommendation would be to open it now

<Yves> if it's just syntax, that's not a major change

Asir: This is not a huge issue - no new feature

<dug> its a bit more than syntax

<dug> he really wants to be on the queue :-)

dug: The usecase is when the metadata is something as simple as wsdl - it is not clear whether this is just a wsld document directly under wsa:metadata or whether it is wrapped in a mex:metadata wrapper. That needs to be cleared up.

Gil: I don't think it's advisable to leave this until last call. We should focus on it now

Asir: Embedding WSDL directly into ws-addressing metadata - was taken out as there was no implementation experience

Ashok: Not relevant here

<dug> I have no idea what WSA 2.2 is

<dug> lol

<asir> Here it is http://www.w3.org/TR/2007/WD-ws-addr-metadata-20070627/#metadatinepr

Bob: I am disappointed that this issue did not get out at the time of issue morotorium

Ashok: Issue 7728 may include this and would be a good location for discussing this issue

<Yves> same as Bob, don't want new issues to postpone LC indefinitely

<scribe> ACTION: Asir to write up proposal with changes incoporated [recorded in http://www.w3.org/2009/12/15-ws-ra-minutes.html#action03]

<trackbot> Created ACTION-130 - Write up proposal with changes incoporated [on Asir Vedamuthu - due 2009-12-22].

<Ram> Resolve with new proposal http://lists.w3.org/Archives/Public/public-ws-resource-access/2009Dec/0045.html

Issue 8200

<Ram> Resolve with new proposal http://lists.w3.org/Archives/Public/public-ws-resource-access/2009Dec/0045.html

Dug: This is about the mex:all dialect uri
... but Gil also notices (8297) has a def of mex dialect from getmedata operation that is wrong

Bob: Any objections to resolving with message 45?


RESOLUTION: 8200 resolved with action described in message 45

Issue 8297

<dug> 8297: http://lists.w3.org/Archives/Public/public-ws-resource-access/2009Dec/0058.html

RESOLUTION: Issue 8297 resolved with proposal in message 58

Issue 8202

<Ram> Proposal: http://lists.w3.org/Archives/Public/public-ws-resource-access/2009Dec/0048.html

<dug> "When this repeating OPTIONAL element is present, the response MUST include only Metadata Sections corresponding to metadata specified by the combination of the URI, Identifier and Content attributes of each of the Dialect elements. For each Dialect element if there is no metadata for that combination of attributes then the response MUST NOT include any Metadata Sections for that Dialect element."

RESOLUTION: Issue 8202 resolved with message 48

Issue 8205

<dug> http://www.w3.org/Bugs/Public/show_bug.cgi?id=8205

Asir: From mailing list - intent is a single element and so we should change schema/text to indicate that it is unique

dug: as there's an xs:any element, we can't prevent multiple mex:md anyhow so there's no point in making it unique

Gil: having individual metadata:section elements in each mex:metadata is no different from multiple metadatasections in one mex:metadata

<gpilz> mex:Metadata(A, B) == mex:Metadata(A) + mex:Metadata(B)

Asir: What we are actually discussing is optimal syntactic form. We should have use case for multiple

Dug: I am happy to write up Gil's issue

<Yves> why can't multiple MEX entries be alternatives and not complement each other?

<Yves> if you have no control, you have no control on the meaning

<asir> agree with Yves

Asir: this doesn't fix it. I would like to understand a motivation. WS-T fixes the extensibility so why can't we

Dug: ws-t isn't the same problem

Gil: There are advantages in having multiple mex:metadata sections in an EPR if multiple components are working on different sections
... I don't understand the motivation of imposing a restriction instead of just having one

Dug: mex:medatadata has extensibility points. Having multiple mex:metadata elements allows folk to have different extensibility elements for each - not restricted to one

Asir: Gil's point is a programming API - a serialisation issue so it's not relevant
... I would like to see a proposal written up with reasons for this function

Yves: If there are multiple mex:metadata elements, then there would be no way of indicating the relationship between the various mex:metadatasections

<asir> Bob: we have to leave for another meeting

bob: this discussion will go to the email list as out of time

<asir> Happy holidays to everyone!!!

<asoldano> +1

<Yves> bye all!

Summary of Action Items

[NEW] ACTION: Asir to write up proposal with changes incoporated [recorded in http://www.w3.org/2009/12/15-ws-ra-minutes.html#action03]
[NEW] ACTION: Gil to produce some specifics for 8284 that reflect the relevant aspects for ws-ra [recorded in http://www.w3.org/2009/12/15-ws-ra-minutes.html#action01]
[NEW] ACTION: Gilbert to produce some specifics for 8284 that reflect the relevant aspects for ws-ra [recorded in http://www.w3.org/2009/12/15-ws-ra-minutes.html#action02]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2010/01/25 00:47:15 $