W3C

- DRAFT -

SOAP-JMS Binding Working Group Teleconference

22 Jun 2010

See also: IRC log

Attendees

Present
Regrets
Chair
Eric
Scribe
Mark

Contents


<trackbot> Date: 22 June 2010

<eric> Mark - you having difficulty dialing in?

yes, Zakim not recognising the tones

trying the US number

better: -)

1) Appointment of the scribe

<scribe> Scribe: Mark

2) Approval of prior meeting minutes

Minutes approved

3) Review the agenda

Mark: Would like to add new issue regarding landing page and broken like

action Eric to fix borken linkon landing page

<trackbot> Created ACTION-181 - Fix borken linkon landing page [on Eric Johnson - due 2010-06-29].

Mark: See http://lists.w3.org/Archives/Public/public-soap-jms/2010Jun/0025.html

4) Review action items

Eric: No progress on actions

Peter: No progress on actions

5) URI specification:

<scribe> No progress

6) Raised spec issues:

<eric> http://www.w3.org/2002/ws/soapjms/tracker/issues/39

Amy: Gives overview of issue

<eric> Text of proposal:

http://lists.w3.org/Archives/Public/public-soap-jms/2010Jun/0023.html

<eric> The original rule : 

<eric>     S MUST copy the JMSMessageID from the original

<eric>     request to the JMSCorrelationID of the response

<eric> Becomes :

<eric>     if there is no JMSCorrelationId set in the request, 

<eric>         S MUST copy the JMSMessageID from the original 

<eric>         request to the JMSCorrelationID of the response.

<eric>     else 

<eric>         S MUST copy the JMSCorrelationID from original 

<eric>         request to the JMSCorrelationID of the response.

Mark: Very much likes the new 2038 protocol proposed by David in http://lists.w3.org/Archives/Public/public-soap-jms/2010Jun/0023.html - suggest we adopt it

Derek: Makes perfect sense, but concerned that this would be a misuse of correlation ID - could we set another message property rather than re-using correlation Id?

Amy: The proposal effectively makes correlation Id into a conversation Id - it is a really clever trick, and it is in the mainstream of JMS to set the Correlation Id first - uses existing technology rather than inventing new headers

Mark: It is a pretty standard technique.. it is used by MQ as a common correlation pattern (COPY CORRELID). JMS makes it more useful because the MsgId can't be set by the sender

Amy: Worried that a new header starts to open the floodgates towards re-implementing WS-Addressing

Peter: Agree - we should not invent new properties, but curious about how existing JMS bridges support correlId. Generally think it is a good idea

DereK: Definitely agree this issue raises a legitimate concern
... and the use case raised by David is a good one

All: No objections to opening this as an issue

Eric: If we specify the use of CorrellationId then it might prohibit using correlation Id for som advanced patterns, but that's OK because WS-Addressing can be used in those cases
... The other option is some other message property for correlation. The disadvantage is that every message has the overhad of another property

Mark: *Could* be done today with a custom property which would appear in the RequestURI rather than inventing a new property (but I still like David's proposal better)

Eric: Could be done with temporary reply Qs

Peter: and this proposal doesn't preclude temporary Qs

Eric: So is David's proposal flexible enough or do we need more flexibility

Amy: Votes for proposal

Mark: +1

Derek: +1 - does not cater for all fringe use-cases but it's a good solution

Peter: +1 Curious to see how JMS bridges would handle it, but this is good

Eric: in favour too - looked back through archives and didn't find anything to conflict with this

Mark: May be better to add to the spec as a separate assertion
... and add new tests

Amy: Will this send us back to last cal?

Eric: No, don't believe so

Amy: If this is important enough to go in, then it must be normative and documented as one of the significant differences between CR and PR

<Yves> agree that it should be documented

Peter: Does this get manifested in any new properties that might be included in WSDL etc.?

Derek: No this is behaviour determined solely by the message sender - i.e. whether to set the correlation id or not

Eric: so we are generally agreed on the proposal. Need to check how much of the binding spec. needs to change.

Action Mark to check how much of the binding spec. needs to change and make a proposal

<trackbot> Created ACTION-182 - Check how much of the binding spec. needs to change and make a proposal [on Mark Phillips - due 2010-06-29].

Eric: We will vote on the proposal next week
... Please discuss concerns or newsuggestion on the list

7) Accepting proposals to close open issues

None

8) Accepting applied resolutions:

http://www.w3.org/2002/ws/soapjms/tracker/issues/38

<scribe> ACTION: Eric to apply resolution to issue 38 [recorded in http://www.w3.org/2010/06/22-soap-jms-minutes.html#action01]

Action Eric to apply resolution to issue 38

<trackbot> Created ACTION-183 - Apply resolution to issue 38 [on Eric Johnson - due 2010-06-29].

<trackbot> Created ACTION-184 - Apply resolution to issue 38 [on Eric Johnson - due 2010-06-29].

9) Moving to PR

<mphillip1> Eric: So any one of us could submit a patch with the latest changes and if a vendor's implementation passes the CXF tests, then we reach the criteria for PR

<mphillip1> Peter: The test validation is based on JAX-WS clients with validation performed by CXF interceptors

<mphillip1> Peter: uses ACtiveMQ but other JMS providers could be plugged in

10) AOB?

<mphillip1> Eric: Should we skip meeting following July 4th?

<mphillip1> All: No vacation plans

Summary of Action Items

[NEW] ACTION: Eric to apply resolution to issue 38 [recorded in http://www.w3.org/2010/06/22-soap-jms-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2010/06/22 16:57:11 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

No ScribeNick specified.  Guessing ScribeNick: mphillip
Found Scribe: Mark

WARNING: No "Present: ... " found!
Possibly Present: All Amy DereK Mark Peter Yves aaaa aabb aacc aadd aaee eric joined mphillip1 peaston soap-jms trackbot
You can indicate people for the Present list like this:
        <dbooth> Present: dbooth jonathan mary
        <dbooth> Present+ amy

Found Date: 22 Jun 2010
Guessing minutes URL: http://www.w3.org/2010/06/22-soap-jms-minutes.html
People with action items: eric

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


[End of scribe.perl diagnostic output]