Minutes WS Choreography WG conference call 10 May 2005
IRC: http://www.w3.org/2005/05/10-ws-chor-irc
Agenda: http://lists.w3.org/Archives/Public/public-ws-chor/2005May/att-0022/Agenda-05102005-0.txt
1. Role Call
------------
Barreto, Chapman, Ross-Talbot, Kavantzas, Tell, Brown, Martin, Lafon
2. Confirm scribe
-----------------
Monica Martin
3. Agenda Changes
-----------------
Add 1108 to agenda.
F2F agenda.
Add exit criteria.
4. Approve minutes
------------------
3rd May
http://lists.w3.org/Archives/Member/member-ws-chor/2005May/att-0004/MeetingMinutes20050503-2.txt
APPROVED
5. Action item review
---------------------
1. ACTION: Martin to do UML diagram from scratch for CDL
IN PROGRESS
Status: No change.
2. ACTION: SRT Check 1027 for issues pertaining to identity
IN PROGRESS
Status: No change.
3. ACTION: Add text in primer or spec to clarify participant relationship/role pertaining to issue 1027
IN PROGRESS
Status: No change.
4. ACTION: Chairs to talk with the XPath 2.0 WG to determine the direction of three-valued logic and existential qualifiers
IN PROGRESS
Status: No change.
5. ACTION: SRT to create a new issue about accessory pertaining to issue 1128. SRT to investigate
IN PROGRESS
Status: No change.
6. ACTION: SRT to rewrite the exit criteria.
IN PROGRESS
SRT: Email sent with G Brown comments from me.
Status: Done.
7. ACTION: chairs to respond to issue raiser for closed issues
STANDING ITEM - Martin will start to close issues this week.
Status: No change.
8. ACTION: Steve to learn about this issue from Nick for the Primer (issue 1079)
IN PROGRESS
Status: No change.
9. ACTION: Yves to define what is meant by correctness
IN PROGRESS
Lafon: There is no definition other than what our working group creates.
We have to create text as a group.
Chapman: Isn't that the action point.
Lafon: I can work on.
SRT: Related to conformance and exit criteria.
10. ACTION: SRT to close issue 1002 and take action to add 1002-addressing
IN PROGRESS
Steve will try to move on some of his actions by next meeting.
Status: No change.
11. ACTION: Charlton, Gary and Nick to discuss 1008 off line to resolve (leaving 1008 on the agenda)
IN PROGRESS
Barreto: Discuss today to reach conclusion if possible.
12. ACTION: Close 971 as Resolved, won't fix
IN PROGRESS.
Status: No change.
13. ACTION: Record 996 as resolved fixed with text in minutes (12 apr)
IN PROGRESS
Status: No change.
14. ACTION: Record 998 as resolved won't fix
IN PROGRESS
Status: No change.
15. ACTION: Record 1018 as resolved fixed
IN PROGRESS
Status: No change.
16. ACTION: Record 1055 as RESOLVED LATER so is editorial
IN PROGRESS
Martin: There was discussion of this on the list this week.
Status: No change.
17. ACTION: Record 1079 as CLOSE WON'T FIX
IN PROGRESS
Status: No change.
18. ACTION: SRT - Put in text related to access and modify in the primer.
IN PROGRESS
Status: No change.
19. ACTION: SRT - Insert 1028 text provided by GBrown into Issue 1128.
IN PROGRESS
Status: No change.
20. ACTION: MC - Close 1110 with the proposed resolution.
IN PROGRESS
Chapman: Want to come back to these in progress issues and actions.
Status: No change.
21. ACTION: Martin to record issue 1102 as resolved/later.
IN PROGRESS
Status: No change.
22. ACTION: Tony to add rns: to the namespace prefix table in the WS-Choreography specification.
IN PROGRESS
Fletcher: Did a sanity check. 'rns' replies to an interface. Another example where the interface
'qname' is tns. Two examples are inconsistent. Email to group whether to change to have both
examples show 'rns.'
Martin: Section 2.5.2.3
Fletcher: Recommend we change this section example to 'rns.' or change all interface examples to 'rns.'
Kavantzas: Prefer to take first approach, 'rns' for interface example in Section 2.5.2.3.
Approved this approach.
Status: IN PROGRESS
NEW ACTION: Fletcher to complete approved approach for #22.
23. ACTION: Steve to close 1128 as resolved will not fix, but add text to the primer about
the use of lists and arrays in CDL.
IN PROGRESS
Status: No change.
24. ACTION: Steve and/or Martin to generate a list of the issue numbers that the editing team
should be working on as a spot check.
IN PROGRESS
Kavantzas: Editors don't have any actions items now.
Status: No change.
25. NEW ACTION: SRT virtual intro Tony to Jonathan Marsh (but to Paul Downey)
Status: Done.
26. NEW ACTION: Tony to mull over and represent next week.
Status: Done (re: conformance).
27. NEW ACTION: SRT to elicit clarification from invited experts
Status: Done.
Chapman: In process of closing issues. Email person of the group decision and notification to issue initiator.
Many issues don't specify resolution. Traceability is very poor.
Kavantzas: Editors' spreadsheet may help.
Chapman: This information needs to be in Bugzilla. I do not wish to do this manually.
Fletcher: Greg could get direction to do this (now or future?).
Chapman: I've done the obvious ones but more time needs to be spent on the remaining ones.
NEW ACTION: Chairs discuss and provide a recommendation.
6. Proposals
------------
1092 http://www.w3.org/Bugs/Public/show_bug.cgi?id=1092
Summary: State relationship between WS-CDL and ebBP
URL: http://lists.w3.org/Archives/Public/public-ws-chor/2005Jan/0034.html
*** PROPOSAL: http://lists.w3.org/Archives/Public/public-ws-chor/2005May/0006.html
15 minutes of discussion
Consensus and/or vote
SRT: Proposal discussed last week and no consensus achieved.
Barreto: Proposal sent out last week and discussion. Have 15 minutes to discuss, and vote if required.
There are two changes as in the proposal.
Tell: There is a problem with ebXML BPSS on agreement of two specifications. CEFACT has an ebXML BPSS.
By June we hope to make progress. Delay to see if agreement.
Discussion item thus far.
Chapman: This may not happen in the current schedule.
We need verbiage for today. Propose CEFACT BPSS. Ask for proof.
Tell: Valid discussion.
Martin: Every spec must have same rigor.
Chapman: We know several of these. There is overlap in message order and sequencing.
Tell: From CEFACT point of view, specifications alike. Specs are more overlapping.
Martin: Are we going to discuss everything from last week?
SRT: We don't argue Java and CDL are competitive to BPEL. This is a tenuous issue in general.
Barreto: BPML and others are mentioned.
SRT: Why not put it in? There is a specific proposal here.
Chapman: I don't know if we prove if they will work.
Martin: The use of 'will' is not RFC 2119 responsibility?
*** PROPOSED AMENDMENT (1): "This specification will work" to "This specification is intended to work".
Chapman: Both CEFACT BPSS and ebBP.
Martin: I would recommend separate proposals.
No objections to AMENDMENT (1) and original proposal.
*** AMENDMENT (1) APPROVED.
*** PROPOSED AMENDMENT (2): Add "CEFACT BPSS" to Section 1.3.
No objections.
Martin: Ask for voice vote.
SRT: Is this an objection.
Martin: Withdraw.
*** AMENDMENT (2) APPROVED.
SRT: CEFACT BPSS
Martin: That schema doesn't validate; how can it be used?
Tell: Relevant to put both with technical specification.
Kavantzas: Is language consistent between computable and executable process languages? Intent was for C#, Java, etc.
SRT: Is this defined in the technical specs referenced?
Martin: Executable language is excluded in our charter.
Kavantzas: They have to be computable semantic definition.
Tell: Why put in if we can't differentiate factors?
Kavantzas: I am questioning this addition.
Martin: Are we going to the original proposal?
SRT: We have consensus on first part (AMENDMENT (1) and AMENDMENT (2), Section 1.3. Issue with Section 1.5 is still in question to computable
semantic definition.
Fletcher: A choreography description language may couple with other languages such those that add further computable semantic definitions.
Replaces "A Choreography Description Language may be couple with other computable semantic definitions, such as those specified in the OASIS ebBP."
*** PROPOSED AMENDMENT (3):
REPLACE
"A Choreography Description Language may be couple with other computable semantic definitions, such as those specified in the OASIS ebBP."
WITH
"A choreography description language may couple with other languages such those that add further computable semantic definitions."
Martin: Propose we amend the original proposal.
*** AMENDMENT (3) APPROVED.
Kavantzas: Don't we need examples in Section 1.5.
Chapman: Editors will do this.
*** SUMMARY OF PROPOSAL to resolve issue 1092 ***
ORIGINAL PROPOSAL: http://lists.w3.org/Archives/Public/public-ws-chor/2005May/0006.html
PROPOSED AMENDMENT (1): "This specification will work" to "This specification is intended to work".
PROPOSED AMENDMENT (2): Add "CEFACT BPSS" to Section 1.3.
PROPOSED AMENDMENT (3):
REPLACE
"A Choreography Description Language may be couple with other computable semantic definitions,
such as those specified in the OASIS ebBP."
WITH
"A choreography description language may couple with other languages such those that add
further computable semantic definitions."
7. Issues requiring clarification
---------------------------------
Summary of progress from last week:
Issues still to resolve:
1008 http://www.w3.org/Bugs/Public/show_bug.cgi?id=1008
Summary: FAULT HANDLING
URL: http://lists.w3.org/Archives/Public/public-ws-chor-comments/2005Jan/0012.html
GB: http://lists.w3.org/Archives/Public/public-ws-chor/2005Feb/0035.html
REMAIN OPEN
Barreto: Only map default to fault name. This is consistent with WSDL v1.1 and v2.0 and Brown agrees.
Add fault name, we can infer fault referenced in both WSDL versions.
Kavantzas: Reusable faults exist in interface level in WSDL v2.0 but happen at choreography level in CDL.
My approach was consistent with this although different. Argument with adding complexity exists with either approach.
Barreto: We wouldn't have this issue with only WSDL v2.0. Fault name provides consistent approach to support
overloaded fault types.
Kavantzas: Your argument doesn't necessitate any change from current approach.
Don't have a consistent way to deal with types between two WSDL versions.
Barreto: Don't have a way to handle overloaded fault types.
Kavantzas: CDL takes responsibility.
Barreto: I don't see how this applies to WSDL v1.1? How can information type support v1.1 WSDL? Information type is associated with a fault type.
Kavantzas: 'may'
SRT: Is this an added capability; does it have benefit?
Kavantzas: CDL already solves the problem. Disambiguation of fault name is in interaction.
Brown: How long will WSDL v1.1 be around? It may be used for some time (see BPEL).
Chapman: That OASIS TC is having same discussion. Don't see reason to change.
Brown: Need a CDL example.
NEW ACTION: Kavantzas provide example showing how WS-CDL binds to faults in WSDL1.1 and WSDL1.2 to show the benefit of his approach.
Nick suggests looking at the example in the 2.5.2.3 Interaction Syntax section: The exchange "badPurchaseOrderAckException"
<informationType name="badPOAckType" type="xsd:string" exceptionType="true"/>
<exchange name="badPurchaseOrderAckException" informationType="badPOAckType" action="respond">
<send variable="cdl:getVariable(,tns:badPurchaseOrderAck,, ,,, ,,)" causeException="true" />
<receive variable="cdl:getVariable("tns:badPurchaseOrderAck", "", "")" causeException="true" />
</exchange>
9. AOB
-------
WAS DEFERRED
MEETING FORMALLY ADJOURNED
After meeting notes
Kavantzas: What about F2F technical issues?
If we don't talk about technical issues, exit criteria, examples, primer and conformance could be discussed.
SRT: We may not make deadline; 1 month delay may take place.
SUMMARY OF OUTSTANDING ACTIONS
-------------------------------
1. ACTION: Martin to do UML diagram from scratch for CDL
IN PROGRESS
Status: No change.
2. ACTION: SRT Check 1027 for issues pertaining to identity
IN PROGRESS
Status: No change.
3. ACTION: Add text in primer or spec to clarify participant relationship/role pertaining to issue 1027
IN PROGRESS
Status: No change.
4. ACTION: Chairs to talk with the XPath 2.0 WG to determine the direction of three-valued logic and existential qualifiers
IN PROGRESS
Status: No change.
5. ACTION: SRT to create a new issue about accessory pertaining to issue 1128. SRT to investigate
IN PROGRESS
Status: No change.
6. ACTION: chairs to respond to issue raiser for closed issues
STANDING ITEM - Martin will start to close issues this week.
Status: No change.
7. ACTION: Steve to learn about this issue from Nick for the Primer (issue 1079)
IN PROGRESS
Status: No change.
8. ACTION: Yves to define what is meant by correctness
IN PROGRESS
9. ACTION: SRT to close issue 1002 and take action to add 1002-addressing
IN PROGRESS
Steve will try to move on some of his actions by next meeting.
Status: No change.
10. ACTION: Charlton, Gary and Nick to discuss 1008 off line to resolve (leaving 1008 on the agenda)
IN PROGRESS
11. ACTION: Close 971 as Resolved, won't fix
IN PROGRESS.
Status: No change.
12. ACTION: Record 996 as resolved fixed with text in minutes (12 apr)
IN PROGRESS
Status: No change.
13. ACTION: Record 998 as resolved won't fix
IN PROGRESS
Status: No change.
14. ACTION: Record 1018 as resolved fixed
IN PROGRESS
Status: No change.
15. ACTION: Record 1055 as RESOLVED LATER so is editorial
IN PROGRESS
Status: No change.
16. ACTION: Record 1079 as CLOSE WON'T FIX
IN PROGRESS
Status: No change.
17. ACTION: SRT - Put in text related to access and modify in the primer.
IN PROGRESS
Status: No change.
18. ACTION: SRT - Insert 1028 text provided by GBrown into Issue 1128.
IN PROGRESS
Status: No change.
19. ACTION: MC - Close 1110 with the proposed resolution.
IN PROGRESS
Chapman: Want to come back to these in progress issues and actions.
Status: No change.
20. ACTION: Martin to record issue 1102 as resolved/later.
IN PROGRESS
Status: No change.
21. ACTION: Tony to add rns: to the namespace prefix table in the WS-Choreography specification.
IN PROGRESS
Status: IN PROGRESS
NEW ACTION (augments 21 above): Fletcher to complete approved approach for #21.
22. ACTION: Steve to close 1128 as resolved will not fix, but add text to the primer about
the use of lists and arrays in CDL.
IN PROGRESS
Status: No change.
23. ACTION: Steve and/or Martin to generate a list of the issue numbers that the editing team
should be working on as a spot check.
IN PROGRESS
Status: No change.
22. NEW ACTION: Chairs discuss and provide a recommendation.
23. NEW ACTION: Kavantzas provide example showing how WS-CDL binds to faults in WSDL1.1 and WSDL1.2 to show the benefit of his approach.
Nick suggests looking at the example in the 2.5.2.3 Interaction Syntax section: The exchange "badPurchaseOrderAckException"
<informationType name="badPOAckType" type="xsd:string" exceptionType="true"/>
<exchange name="badPurchaseOrderAckException" informationType="badPOAckType" action="respond">
<send variable="cdl:getVariable(,tns:badPurchaseOrderAck,, ,,, ,,)" causeException="true" />
<receive variable="cdl:getVariable("tns:badPurchaseOrderAck", "", "")" causeException="true" />
</exchange>
-- discuss next week