Minutes of W3C WS Choreography WG conference call held on 17th February 2004,1pm PDT


17 February 2004




 Martin Chapman


Steve Ross-Talbot




W3C Staff Contacts


Carine Bournez


Yves Lafon





Anthony Fletcher

Choreology Ltd

David Burdett

Commerce One

Yoko Seki

Hitachi, Ltd.

Assaf Arkin

Intalio Inc.

Abbie Barbir

Nortel Networks

Nickolas Kavantzas

Oracle Corporation

Ivana Trickovic


Ugo Corda

SeeBeyond Technology Corporation

Monica Martin

Sun Microsystems, Inc.


  Raw IRC log at:: http://www.w3.org/2004/02/17-ws-chor-irc


Appointment of scribe:


Assaf Arkin from Intalio volunteered to scribe.


The following is a listof recent scribes (in order): Tony Fletcher, David Burdett, Tony Fletcher, Monica Martin, Ugo Corda, Mayilraj Krishnan, Ravi Byakod , Martin Chapman, Steve Ross-Talbot, Monica Martin, Nick Kavantzas, Ed Peters, Anthony Fletcher, Jeff Michkinsky,  Dinesh Shahane, Greg Ritzinger, Ed Peters, David Burdett,Ivana Trickovic, Ugo Corda, Assaf Arkin, Monica Martin, Carol McDonald, Nick Kavantzas, Tony Fletcher, Mayilraj Krishnan, Francis McCabe, Jeff Mischkinsky, David Burdett, John Dart, Monica Martin,Tony Fletcher, Jim Hendler, Kevin Liu, TonyFletcher, Jon Dart, DavidBurdett,Ed Peters, Greg Ritzinger, Monica Martin, Len Greski, Jean-JacquesDubray,Monica Martin, Mayilraj Krishnan, Francis McCabe, Michael Champion,AbbieBarbir, David Burdett, Jon Dart, Carol McDonald,Yaron Goland, Leon Greski,Ed Peters, Greg Ritzinger, Daniel Austin, PeterFurniss, Jim Hendler


Approve minutes

Due to a technical error there are no formal minutes of the meeting held on 10 February 2004.

Tony wrote some notes of the meeting,  but the scribes notes are not available. Tony gets credit on the scribe list!


Action item review

Because of the minutes will review the action items frm 10 Feb 2004


ACTION: chairs should also ask marco to see if his requirements are captured within the latest req-doc (0113, 0120, 0127, 0203, 0210, 0217) NO PROGRESS.


SRT: action 1 no progress


ACTION: SRT and MM will provide further detail on the “banana-calculus” discussions in order to record this exception/errors discussion properly.  (0127, 0203, 0210, 0217) NO PROGRESS.


 SRT: action 2 no visible progress


ACTION:SRT will update bugzilla entries to reflect exceptions/error discussion. (0127, 0203, 0210, 0217) NO PROGRESS.


SRT: action 3 still no progress (depends on previous)


ACTION: Nick to define the features required of an intermediate end-point language (0113, 0120, 0127, 0203, 0210, 0217) NO PROGRESS.


SRT: action 4 will spill into F2F, requires discussion with Nick


ACTION: convert ER diagrams into UML class diagrams (0113, 0120, 0127, 0203, 0210) DONE.


SRT: action 5 (ER UML diagram) been completed during the course of the call

DB: I sent it in PDF



ACTION:editors to revise documents and table of contents according to comments received (0203, 0210) DONE


SRT: action 6 is done


ACTION: Chairs to look into a far east meeting in sep/oct (0113, 0120, 0127, 0203, 0210, 0217) IN PROGRESS.


Yves will chase at Plenary

SRT:also checked with mayor of Cape Town to see if facilities are available, no response yet


ACTION: Chairs to organise an issues conference call in which the regular conf call is used with members to look at the issue in bugzilla (possibly the Jan6th conf call) (0113, 0120, 0127, 0203, 0210, 0217) NO PROGRESS.


SRT: action 8, no progress


ACTION: Chairs  to raise a topic of Binding, Context etc. (0113, 0120, 0127, 0203, 0210, 0217) IN PROGRESS.


SRT: action 9, no progress, still to be done by the chairs


ACTION: Editors to issue requirements document for review on public mailing list for next call (0127, 0203, 0210) DONE.


SRT: action 10 has been done, many thanks to Abbie for delivering the document, and Daniel for his help


ACTION: Chairs to schedule conference call to review requirements document (see above)(0127, 0203, 0210, 0217) NO PROGRESS.


SRT: action 11, now that we have requirement document, will schedule appropriate time for call to discuss it

ACTION: Steve/Martin - let us see some email traffic to manage the process to get the requirements doc reviewed quickly and published -if we cannot make it by 13th (which is tight) then we inform the members as such. (0106, 0113, 0120, 0127, 0203, 0210, 0217) DEPENDENT ON REQ DOC REVIEW (see above)

SRT: action 12, we would like people to look at this requirement and start a thread this week, appreciate if people can look at use cases vs requirements and see if there are any gaps


ACTION: Steve to ask WSD WG whether they intend to support publish/subscribe MEP. (0120, 0127, 0203, 0210, 0217) DONE



ACTION: Team to review the Model Overview prior to next meeting. (0127, 0203, 0210, 0217) IN PROGRESS.



ACTION:   Greg to log Monica's issues into Bugzilla. (0203, 0210) DONE


 DB: Greg said it's done


Tracking Items:


SRT: filling in for Abbie: requirements document was sent out. we have a document we can review. Martin and I will be meeting on Friday to do a fine comb review of the document and report back.

Specification Editors Report

DB: no further progress on spec.

Issues Discussion:


SRT: issues tracking and resolution status: no major issues logged since the last call


BPSS Monica: had face to face, outlining scope for next version, some of the pieces discussed in face to face will have a draft schema coming up next month


BPEL: nothingto report. Request for feedback still outstanding.

NEW ACTION: MC, check out with BPEL chairs tomorrow on a feedback report 


Plenary F2F Agenda


draft f2f agenda http://lists.w3.org/Archives/Member/member-ws-chor/2004Feb/att-0013/Mar_04_F2F_Agenda_-_0.txt


MC: action to check out with BPEL chairs tomorrow


SRT: I'd like to go through one of the use cases and all its variants and at least form an abstract sense map a solution that meets those requirements with respect to the model overview


SRT checked with mayor of Cape Town to see if facilities are available, no response yet

SRT: quality check to make sure we're meeting the requirements we laid out in the document

 MC: the only danger when we start modeling and shift the requirements document around

DB: easier take requirements, check if model supports them; since model is model, hard to see what it looks like

SRT: I buy this argument; the requirement document may change over time

DB: get more agreement about the model, take the model and see what it looks like as XML, might be a bit too early

Tony: I agree with David, we need to cleanup and sign on the requirement document, then do the model overview and go to the language from that, then we can provide examples for one or more use cases and close the loop

SRT: I withdraw my request

Yves: on the logistics page there should be indication for social activities, there's reception on first day, nothing else planned right now

SRT: hot topics, propose to take one of those topics and then leave 5 mins for other business


Hot Topics



SRT: hot topic: WSDL pub/sub

DB: pub/sub is the idea you ask service to tell you when they do something, and then they tell you; subscribe to information feed

Pub/Sub WSDL came about from this response.


DB: various suggestions on how to make it work, one being choreography; various ways to do it

Monica: I agree, we need to put boundaries around it

DB: need to pass a delivery address to where notifications get sent; that would be common sequence of messages; when you get down to the details you find differences in how you do it, ways you may want to do it; for example, if I don't think ebXML reg/rep is best for auction, or if I want to use ebXML reg/rep but add credit validation, is that something ebXML reg/rep is designed to do? might gut feeling is that it's not designed to do that

DB: should pub/sub be passive and have a standard way to do it that people could reuse (in WSDL) or should we define it as choreography and let WSDL be as it is; right now exploring whether it should be a WSDL atom, or rely on existing WSDL atoms

SRT: you can certainly describe pub/sub pattern without knowing how many participants there are, it's replication like pi-calculus, and you can hide certain behavior so system appears like pure pub/sub; if you've used real pub/sub products there's always a broker inbetween, none of them are true pub/sub; the issue gets tricky in WSDL if there's a pub/sub mechanism underneath; maybe it's not possible to model a true pub/sub in that sense

Monica: you may not know number of participants, but may define boundaries on how you trigger certain events and interactions around pub/sub pattern, e.g. conversation, some things not triggered by timeout, may require certain number of participants to trigger a certain event

DB: like a car pool, you trigger a car pool when you get 8 people together

Monica: event notification could be possible to trigger those events that could result in message exchange

SRT: is event notification same as notification pattern in WSDL?

MC: probably not, all it is is an out pattern; in pub/sub you need cardinality, whether it's one person or multiple persons, and that's missing in WSDL

MC: choreography should be able to perfectly model a pub/sub system talking to a broker

MC: that's all about putting MEPs together

SRT: that might be enough given that the charter drives us towards WSDL 2.0, it will give us a level of utility above that

MC: we need proof of concept that we can model this using choreography

Tony: at best it's split responsibility, some support given in WSDL, but WSDL wouldn't do everything; there's more than one pattern, more variations for what would happen, and you should be able to express those in choreography, you don't want to build them all into the infrastructure; and to make it work, you need a sort of business level semantics you don't get in WSDL; a single system may have several subscriptions with a service and when it gets a notification

DB: there's some mileage to have a standard pub/sub choreography that people could use and reuse; certain groups of choreography would then become part of the infrastructure and then you can use them at the higher level

Monica: need to evaluate what's in our scope



SRT/MC: preference not to meet next week, have a breather before the F2F, use time to read through documents and get up to speed

SRT: main homework: req document, model overview with new diagram


Yves: not sure about dail-in capability, will check that


Summary of New Actions

ACTION:  Greg to log the number of levels defined in the in language


ACTION: MC, check out with BPEL chairs tomorrow on a feedback report