|
Chairs: |
|
|
Oracle | |
|
Enigmatec | |
|
|
|
|
W3C Staff Contacts |
|
|
|
|
Attendees:
|
BEA Systems | |
|
Cisco Systems Inc | |
|
Intalio Inc. | |
|
Nortel Networks | |
|
Novell | |
|
Oracle Corporation | |
|
SAP AG | |
|
SAP AG | |
|
SeeBeyond Technology Corporation | |
|
Software AG | |
|
Sun Microsystems, Inc. | |
|
TIBCO Software | |
|
W. W. Grainger, Inc. | |
|
webMethods, Inc. | |
|
BEA Systems |
irc log at http://www.w3.org/2003/10/28-ws-chor-irc
Ivana Trickovic (SAP) scribed.
The following is a list of recent scribes (inorder): Ugo Corda, Assaf Arkin, Monica Martin, Carol McDonald, Nick Kavantzas, Tony Fletcher, MayilrajKrishnan, Francis McCabe, Jeff Mischkinsky, David Burdett ,John Dart, Monica Martin,Tony, Fletcher, Jim Hendler, Kevin Liu, Tony Fletcher, Jon Dart, David Burdett,Ed Peters, Greg Ritzinger, Monica Martin, Len Greski, Jean-Jacques Dubray,Monica Martin, Mayilraj Krishnan, Francis McCabe, Michael Champion, AbbieBarbir, David Burdett, Jon Dart, Carol McDonald, Yaron Goland, Leonard Greski,Ed Peters, Greg Ritzinger, Daniel Austin, Peter Furniss, Jim Hendler
Minutes 21st October 2003 (change URI)
Minutes were approved
Actions from previous meeting marked with *
Organizational
ACTION : The chairs are seeking editors for the main specification document. Volunteers should signal their availability to the chairs via email. Volunteers will be selected next week. (IN PROGRESS, 1021, 1028)
ACTION : chairs look at WSA issues process and recommend whether it should be adopted by this group. ( IN PROGRESS,1021, 1028)
Co-chairs are highly involved in the upcoming editors f2f meeting. There was a suggestion to cancel the next conference call scheduled for Tuesday, Nov 4. There were no objections. The next W3C WS-Choreography WG conference call will be held on Tue, Nov 11, 2003 at 1pm PST.
Martin will move this week to Dublin, Ireland. The email address will remain the same.
Use cases/requirements
ACTION: Use Case proposers to highlight/extract any EAI specific aspects from their use cases. (DONE)
ACTION: Chairs to reply to Marco requesting clarification of his use case. ( IN PROGRESS, 1021, 1028)
ACTION: Steve: Will put all the comments on draft requirements above into the requirements spreadsheet and send out. ( NO PROGRESS, 1021, 1028)
SteveRT: Daniel sent the spreadsheet, but there is no progress.
ACTION: Editors of the requirements are directed to look at the issues list and filter each issue in a similar way to the filtering method used at the F2F. ( IN PROGRESS, 1021, 1028) – to be taken into account at editors meeting in November.
SteveRT: Editors of the requirements document will do this next week, during the editors f2f meeting in UK.
Issues
ACTION : SRT Brought semantics question to the TAG. On chairs coordination call, he asked about semantics for/of choreography. A new Semantic Web Services Interest Group is being formed in about one month. Issue will be sent to that group when it is formed. (NO PROGRESS, 1021, 1028)
ACTION: Steve RT will send a one-page summary of his thoughts ( NO PROGRESS, 1021, 1028)
5. Standing tracking items (a section designed to ensure that longer running items are properly tracked)
ACTION: Daniel to look though document and see which requirements we captured so far regarding transactions. (NO PROGRESS,1021, IN PROGRESS 1028) http://www.w3.org/2003/10/14-ws-chor-irc- T21-05-54
SteveRT: Asked to be done before the upcoming editors f2f meeting.
ACTION: Tony to take a look at Oracle submission and see if it meets base requirements (DONE) http://www.w3.org/2003/10/14-ws-chor-irc- T21-06-11
Tony saw that it meets the requirements, but argues with the way it is currently done. Further discussions are required.
*ACTION: (Chairs) Further discussion needs to be on agenda. (INPROGRESS, 1028)
*ACTION: (Tony) Explain (in email) the substance of the issue.
Status:(IN PROGRESS, 1028)
SteveRT: Tony is not on the call. Tony should explain why this is an issue. Chairs are waiting for the explanation, and then will put further discussion on agenda for one of the upcoming conference calls.
No major issues to report.
David Burdett is not on the call. We cannot discuss requirements matching.
Steve put 5 items that had been derived from email discussions on the agenda.
SteveRT: Tony Fletcher has raised this issue. Tony is not on the call. Defer the discussion.
NEWACTION ITEM: Co-chairs bring Transaction issue on agenda for the next conference call.
SteveRT: I am referring to figure included in section 1.2 – Choreography of Action
Nick: It is a useful diagram.
Nick also sent an email with comments on Steve’s feedback.
Steve: WS-CDL has more passive approach.
Nick: WS-CDL is about defining common observable behavior; it can be used to generate behavioral interfaces, abstract processes - the behavior from the point of view of a single participant.
Nick: Brief explanation of the notation:
· dots are channels
· interaction is moving from one role to another; different documents are exchanged between roles
· reactions show dependencies between interactions; reactions can be mutually exclusive
· many reactions can be defined for a single state
…
Discussion: should the model of choreography assume a central control place?
Nick: Believes that a central control is not possible
Daniel: Agrees that a central control must not be requested. Is there any explicit requirement on the central control? We must be sure that this is included in the requirements document. The editors of the requirements document should find appropriate wording for that.
MikeChampion: Is this the main differentiation between WS-CDL and BPEL?
Yaron: BPEL does not require any central control. Also, what we [W3C WS Choreography Working Group] are doing can also be done by BPEL; the way to do that in BPEL is to use abstract processes.
Two problems are identified:
· Conformance; can we check whether abstract and executable parts are aligned
· Compatibility; can we check whether two abstract processes are mirror images of each other
Nick: Refers to his discussion with Toni during the last BPEL TC f2f meeting; it has been questioned whether it is easy to find mirror images of abstract processes, especially if there are more then two parties
Yaron: It is possible to do mirroring; technically it could be possible but for a human being it will be proved to be impossible
Yaron: This group should act more actively and explain the scope
Daniel: Is there any concrete suggestion?
Yaron: A concrete example should be provided.
Martin: When we get the language we can do that; now it is too premature to show anything
Yaron: Does not agree; it is premature
SteveRT: Scalability issue – building distributed systems on Web already discussed last time; we have to explain the business benefit; we do not have a concrete use case to show differences between BPEL abstract processes and WS-CDL; it would be good to have that
SteveRT: how can we differentiate between BPEL abstract processes and WS-CDL?
Can we draw any parallel with the differentiation between message-oriented middleware and spaghetti connections? Can this be used as an argument for using WS-CDL instead of BPEL?
Yaron: It is hard to position this group.
Martin: When we get a baseline document we can do that.
SteveRT: I am going to write a paper on this (should include a use case); the paper should try to position the choreography language and will address businesspeople; any input is welcome.
Steve What is the meaning of Recovery? What does it mean for choreography?
Nick: there are two things:
· backward recovery
· forward recovery
Nick continues the explanation: We should be able to recognize that something went wrong. The language allows us to specify conditions (exceptional conditions).It does not mean there is a change of state of the participant. For choreography is this important to have condition. We are talking about the state of choreography and not private state of any participant.
Nick: WS-CSL allows using information exchanged in any previous interaction.
Yaron: Choreography does not have to capture all information.
Nick: In some cases it is needed to have information that is exchanged previously and use it in order to control the flow of subsequent messages.
Yaron: The set of features should be restricted. Argues that those kinds of dependencies are not needed. They are saying too much; the language should be simple.
Nick: Which features are not needed in v1.0, but can be included in v2.0; how far we would like to go?
Yaron: We do not need pre-conditions and post-conditions.
Nick: Then we have a linear model.
Yaron: We will still have fork/join, etc.; conditions do not have to be computer processable.
Nick: There are going to be a modeling framework producing WS-CDL and BPEL.
Yaron: CDL will be a persistence format for a tool that people will use in the negotiation phase; with less information, we get loosely coupled systems.
Nick: CDL must be used to produce abstract processes (this is what we need at the end)
SteveRT: Regarding message structure mismatch; are you [Nick] proposing that WS-CDL should do transformation?
Nick: No
SteveRT: Condition information must be somewhere.
Yaron: Too many things will be exposed in the global model.
Nick: There is a problem to define rules for creating mirror images of abstract process(es) – this is a technology failure
Daniel: Yaron’s proposal for a minimalist approach does not work because this kind of proposal will not meet all requirements of the group.
SteveRT: How can we resolve this? An example from Nick with pre-conditions and post-conditions and guards could help.
Daniel: If we try to make such a change I would like to see a good motivation.
Martin: The requirements document must be reviewed.
Monica: We should prioritize what must be in the shell; how to set priorities?
Tryt o be pragmatic. Also know what we can do.
Martin: This will be included with CSF.
Steve: I strongly suggest that members should review the requirements document and ask questions.
SteveRT: Asks Nick to summarize his comments in
an email.
NEW ACTION ITEM: Nick to carlify what is meant by Semantics
of Web Services by email.
EAI use case: will be discussed next week.
NEWACTION ITEM: Co-chairs bring EAI use case for the next conference call.
No other issues
ACTION: Requirements editors to segment CR005 into a local exception handling case and a cross domain case. (STATUS???)
ACTION : The chairs are seeking editors for the main specification document. Volunteers should signal their availability to the chairs via email. Volunteers will be selected next week. (IN PROGRESS, 1021, 1028)
ACTION : chairs look at WSA issues process and recommend whether it should be adopted by this group. (IN PROGRESS,1021, 1028)
ACTION: Chairs to reply to Marco requesting clarification of his use case. ( IN PROGRESS, 1021, 1028)
ACTION: Steve: Will put all the comments on draft requirements above into the requirements spreadsheet and send out. (NOPROGRESS, 1021, 1028)
ACTION: Editors of the requirements are directed to look at the issues list and filter each issue in a similar way to the filtering method used at the F2F. ( IN PROGRESS, 1021, 1028) – to be taken into account at editors meeting in November.
ACTION : SRT Brought semantics question to the TAG. On chairs coordination call, he asked about semantics for/of choreography. A new Semantic Web Services Interest Group is being formed in about one month. Issue will be sent to that group when it is formed. (NO PROGRESS, 1021, 1028)
ACTION: Steve RT will send a one-page summary of his thoughts ( NO PROGRESS, 1021, 1028)
ACTION: Daniel to look though document and see which requirements we captured so far regarding transactions. (NO PROGRESS,1021, IN PROGRESS 1028) http://www.w3.org/2003/10/14-ws-chor-irc- T21-05-54
ACTION: (Chairs) Further discussion needs to be on agenda. (INPROGRESS, 1028)
ACTION: (Tony) Explain (in email) the substance of the issue.
Status:(IN PROGRESS, 1028)
NEW ACTION ITEM: Nick to clarify what is meant by Semantics
of Web Services by email.
NEWACTION ITEM: Co-chairs bring EAI use case for the next conference call.