|
Chairs: |
|
|
Oracle Corporation | |
|
Enigmatec Corporation | |
|
W3C Staff Contacts |
|
|
BEA Systems | |
|
Choreology Ltd | |
|
Cisco Systems Inc | |
|
Commerce One | |
|
Computer Associates | |
|
Enigmatec Corporation | |
|
Fujitsu Ltd | |
|
Hitachi, Ltd. | |
|
Intalio Inc. | |
|
Intalio Inc. | |
|
National Computerization Agency | |
|
Nortel Networks | |
|
Novell | |
|
Oracle Corporation | |
|
Oracle Corporation | |
|
SAP AG | |
|
SeeBeyond Technology Corporation | |
|
Sun Microsystems, Inc. | |
|
TIBCO Software | |
|
W. W. Grainger, Inc. | |
|
webMethods, Inc. |
|
W. W. Grainger, Inc. | |
|
Sun Microsystems, Inc. |
Mayilraj Krishnan, Francis McCabe, Michael Champion, Abbie Barbir, David Burdett, Jon Dart, Carol McDonald, Yaron Goland, Leonard Greski, Ed Peters, Greg Ritzinger, Daniel Austin, Peter Furniss, Jim Hendler
Monica Martin confirmed as scribe
No issues were raised. Passed
ACTION: Daniel to post logistics information. Done
ACTION: Jon Dart. Summarize use cases from privacy discussions. – Done, see mailing list
NEW ACTION:
DA will own the privacy discussion thread and summarise the requirements that ensue. http://lists.w3.org/Archives/Public/public-ws-chor/2003May/0167.html
ACTION: Carol to look into WSCI summary. Carol not on call – No progress
ACTION: SRT to capture reqs from minutes. No progress
ACTION: All: alternate formal models other than the pi-calculus. Agenda item for next F2F – proposal from FM Event Calculus) In progress
ACTION: Yves to setup reading list page. SRT will work with Yves. No progress
ACTION: DA will do the wordsmith, summarize and put together revised mission statement. Once again it is not final. (see agenda item)
ACTION: SRT to extract summaries of harvested specs (wsci, bpel). No progress
ACTION: Co-chairs to harvest agenda items for F2F
SRT: We need to harvest agenda items for F2F. Reported that Tim Berners-Lee to F2F which emphasizes the importance the work with the OASIS liason.
NEW ACTION:
Grainger to ask for repost on F2F logistics.
MC: Opportunities for this group?
JD: Reporting on feedback from Tibco representative reported: Confusion about the process was unavoidable given the number of people. Not a lot of substantive work was done in F2F. Submitters have lengthy issues reportedly but they have not been given to the TC membership. Concerns continue with the BPEL IPR submission. Request has been made for the submitters to renounce IP claims, or to publish the licenses under which the spec will be made available.
YG: BEA is just getting their terms ready for publication. See BEA site and search for BPEL for generic license.
MC: OASIS lacks rules on how to manage the TC.
MC/YG: Significant number of people were reading the specification in the meeting.
YG: Open issues list posted on members site. Little feedback due to lack of specification understanding.
MC: Wasn't it true that the issues raised were put in the parking lot? Silent consensus.
MC: Technical details were acknowledged but not addressed; scope was severely managed. BPEL ready to co-locate of two F2F with WS-Chor and OASIS BPEL.
YG: Need formal process to provide issues - Dieter Roller, Jeff Mischkinsky, and Yaron Goland. Group decide the work item. Chairs can not rule items out of scope.
MC: Meeting was closed on taking issues.
JD: Too early to understand how substantive changes will be.
YG: Now about 50 people present and maintaining membership.
MC: The TC lead leverages the policy and can apply discretion.
YG: Indicated Karl Best verified that the TC lead does have discretion, but the guidelines may be strictly applied by this group.
???: What about logistics between two TCs?
DB: Domains of control and contract - overlap with abstract process definitions within BPEL.
YG: Agrees. MS and IBM strongly believe in abstract processes and are concerned about the overlap with WS-Chor.
DB: Seem to be focusing on that area in WS-Choreography.
YG: BPEL abstract process is a program for particular process. The global view is for each participant requires a unique BPEL abstract process. Then, invent a global plug map so say which processes work together. You could then reverse engineer a global model, and have artifacts unique to a particular participant.
DB: This creates problems for BPEL and perhaps opportunity for WS-Choreography.
YG: Introduce plug maps into BPEL then using abstract processes you could do every possible process. But this is not necessarily a clean way to generate a global view.
DB: WS-Chor specify what they think their opportunities are.
SRT: Continue conversation on member list only.
DA: Goland proposal is interesting. Hopefully come to closure at F2F.
MC/SRT: Don't divulge how we are working with or around BPEL on the list.
SRT: Do what we are going to do.
JD: Technical conversations should be on the public list.
DA: Let's not plan on dependencies on BPEL. Think about what Goland has said about the pi-calculus mapping.
SRT: Greg Meredith paper was provided to members list and what is meant by contracts. This is an opportunity for WS-Choreography.
TF: Part of BPEL meeting was superficial as people met one another. Submitters were trying to impress upon TC their intended scope. BPEL is single role looking out. The global model has some overlap to this view (role abstract process). At least two roads exist for WS-Chor.
JM: Our activities are not invisible so we have to lay a plan and do it. We are working in public and the other TC is 'sort of working in public.' Do the right thing.
JJD: If there is no coherent work between the groups, it will be impossible to align the semantics.
MC: That requires both groups to work together. Dual team members have to make sure it works not particularly the liaisons.
SRT: Has a problem with abstract problems being licensed in this way. Joint membership can drive towards complementing as opposed to forcing competition.
MC: WSDL support and other constructs should be done appropriately.
DA: The mission of the WS-Chor WG at W3C is to specify one or more XML-based languages building on the foundation of WSDL 1.2 to provide interoperability among Web Services in their communications and ordering of events.
The mission of the WS-Chor WG at W3C is to specify the means by which Web Services interoperate how composition of Web Services is performed, and how the sequencing of events among services may be regulated to ensure interoperability.
YG: Objects to the use of 'language.' BPEL misinterpretation could occur. This is a political modification.
SRT: What about ‘description’ instead of ‘language’?
YG: Yes ‘description’ would work for me.
JD: Can we combine two statement proposals. Group is restricting realization of this goal, based on the first proposal.
FM: One difference is reference to composition in the second proposal. First proposal doesn't preclude composition and language to that effect could be added to it.
JM: Challenges Goland on the focus of the word ‘language’. Are we in competition with XML or WSDL working groups given they also use the word ‘language’? He disagrees and said they could be complementary.
YG: This is a unique circumstances and sensitivities.
MC: How much do we have to do to accommodate competition and perceptions with BPEL?
DB: Shares concerns with Goland. Filter message appropriately - be proactive to alleviate confusion.
SRT: Analysts discussions at Budapest and all around - there is a great deal of confusion. WS-Choreography is WSCI in the press.
DA: (on IRC) We can't avoid conflict with BPEL.
DB: Get out a clear message and BPEL complements where appropriate.
MC: Mission statements are too broad.
JJD: WS-Chor can add value if keep independent from company boundaries. Similar patterns between components exist. Huge value if we can describe choreography that accommodates internal and external.
DB: What about domains of control, not between businesses?
JD: This is not addressed in mission statement.
SRT: We have no alternative but to place the mission statement on the mailing list for further discussion.
MC: How do we express decisions in a choreography?
DA: Is the decision point in the choreography? A possible way to solve problems has been proposed, although he is not sure if it is necessary.
DB: Decision point doesn't influence message flows. Don't have to document all the decision points. All you need to know are the outcomes unless an entity needs to communicate the reasons for the outcomes.
FM: Reinforced Burdett point. You can do validation of the choreography unless you are capable of exposing decision points.
YG: Agrees. The issue is not whether or not you expose them, but do you expose the logic used for the decision point. BPEL does this.
JJD: This is an architecture problem - client-server vs. peer-to-peer – you need to understand if the message makes sense. Need tools to do that for the contract. The entity will not have point of reference on the message received. If you move this to the orchestration, this is not workable. Business Service Interface does this.
MC: Supports Burdett points. Is the decision point computationally complete?
FM: Finds that last point as a red-herring. The issue is public or private state. How can we express states? Consensus is not to expose private state; expose public state.
JJD: Complex problem.
DB: All these ideas need to be discussed with BPEL.
MC: (on IRC): Why don't we talk shared state, i.e. the collection of each participant's public state.
DA: (on IRC) said this was 'system state.'
JM: Confused about private and public state? Do we expose state – and by definition it is public state.
JJD: Problem is the sychronization. Message exchange level and a second level where you assign meaning. What state are we talking about - the state of what?
FM: Social factor of the state
TF: Confused by the discussion. Want to be able to express choreographies that are non-deterministic. To Dubray's point, you need to sychronize the states. Need to express the decision point and some logic. The instance may have restictions that makes it more deterministic. Should be flexible on how and what is expressed.
???: Partial executability thread: Send summary to the Chapman and SRT first.
ACTION: Jon Dart. Summarize use cases from privacy discussions. – Done, see mailing list
ACTION: Carol to look into WSCI summary. Carol not on call – No progress
ACTION: SRT to capture reqs from minutes. No progress
ACTION: All: alternate formal models other than the pi-calculus. Agenda item for next F2F – proposal from FM Event Calculus) In progress
ACTION: Yves to setup reading list page. SRT will work with Yves. No progress
ACTION: DA will do the wordsmith, summarize and put together revised mission statement. Once again it is not final. (see agenda item)
ACTION: SRT to extract summaries of harvested specs (wsci, bpel). No progress
ACTION: Co-chairs to harvest agenda items for F2F
ACTION: DA will own the privacy discussion thread and summarise the requirements that ensue. http://lists.w3.org/Archives/Public/public-ws-chor/2003May/0167.html
ACTION: Grainger to ask for repost on F2F logistics.