|
Chairs: |
|
|
Oracle | |
|
Enigmatec | |
|
W3C Staff Contacts |
|
Attendees:
|
Name |
Company |
|
Choreology Ltd | |
|
Cisco Systems Inc | |
|
Commerce One | |
|
National Computerization Agency | |
|
Nortel Networks | |
|
Novell | |
|
Oracle Corporation | |
|
SAP AG | |
|
Software AG | |
|
Sun Microsystems, Inc. | |
|
TIBCO Software | |
|
webMethods, Inc. |
irc log at ?
Ed Peters (webMethods) volunteered to scribe.
The following is a list of recent scribes (inorder): David Burdett, Ivana Trickovic, 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 11th November 2003
Minutes were approved
ACTION: PLEASE Let SRT know if they are planning to attend the F2F as they need to know the accommodation in the college.
SRT: venue IS booked, registration page should be up soon.
(IN PROGRESS, 1111, 1118)
ACTION: Steve/Yves – set up a registration page for the December FTF (IN PROGRESS, 1111, 1118)
ACTION: Requirements editors to segment CR005 into a local exception handling case and a cross domain case.
Action Item not understood. If no clarification id forthcoming it will be closed at next meeting. (DELETED)
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. (DONE)
(srt) editors are communicating among themselves
(mc) normal practice is to have a separate editors list
(srt) once editors have met, they need to publicly state what they're doing
ACTION : chairs look at WSA issues process and recommend whether it should be adopted by this group. (IN PROGRESS, 1021, 1028, 1111, 1118)
ACTION: Chairs to reply to Marco requesting clarification of his use case. (DONE No reply as yet)
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. To be taken into account at editors meeting in November. ( IN PROGRESS, 1021, 1028,1111, 1118)
(srt) editors have been too busy trying to put together new version of req. doc.
(srt) Abbie is editing for an ailing Daniel for the meantime
ACTION : SRT Brought semantics question to the TAG. On chairs coordination call, he asked about semantics for/of choreography. A new SemanticWeb Services Interest Group is being formed in about one month. Issue will be sent to that group when it is formed. (NO PROGRESS, 1021, 1028,1111, 1118)
ACTION: Steve RT will send a one-page summary of his thoughts ( NO PROGRESS, 1021, 1028, 1111, 1118)
NOTE THESE ARE THE SAME ACTION ITEM.
5. Standing tracking items (a section designed to ensure that longer running items are properly tracked)
Requirements next steps (progress/review)
ACTION: Daniel to look though document and see which requirements we captured so far regarding transactions. (NO PROGRESS, 1021, 1028, 1111, 1118)
ACTION: (Tony) Explain (in email) the substance of the issue (on transactions).
Status: (IN PROGRESS, 1028, 1111, 1118)
(tony) In reading the minutes, the context of comments requested from me was unclear. still waiting to get some clarification on what that is
(srt) substance of issue:
(srt) (1) you accepted that Oracle submission met your requirements, but took issue with the way it did so -- please explain why.
(srt) I wanted this to be done in email because it had been discussed in great detail on conference calls
Requirements document review
(srt) standing item: document status / use case feedback
(srt) open this up to group about work on requirements documents
(srt) BTW, Tony has also been supporting the work on the requirements document
(tony) document moves us on quite a lot, like the new structure
(tony) beginning and introduction are good, use cases work well to illustrate requirements
(tony) the backend of the document has the feel of a dumping ground, and sections needing to be filled out
(mc) can we get a rationale for the requirements section itself?
(srt) this is section six
(mc) is this essentially a cut-and-paste of the spreadsheet?
(ab) this table will be cleaned up based on the work from the face-to-face
(ab) duplicate and deleted items were left in so people can see what we're doing
(mc) this should be cleaned up, because it can definitely be confusing
(ab) so we'll remove deleted/duplicate items and renumber
(srt) two bits of work to be done:
(srt) (1) sweep through spreadsheet and make sure we delete things that need deleting
(srt) (2) we were given loose direction to reclassify some requirements
(ab) I want this version uploaded to the server somewhere for comparison if needed
(ab) Then we delete everything that needs deleting, and renumber if needed
(ab) so this section was left as is on purpose, to have a reference point
(ab) we'll reissue it late this week / early next week, cleaned up
(srt) any other comments from membership?
(mm) are you willing to take comments before week is over, or do you want them now?
(srt) I'd prefer to get them out in the open as soon as possible, and discussion helps
(mm) I sent in a comment about the front-end of the document, and how we use the word "contract"
(mm) we don't distinguish technical vs. business contracts, and we should do that
(mc) that distinction should be made up-front, because we're definitely not doing business contracts
(srt) that's my action
NEW ACTION: SRT to edit section 2.1-2.3 to make clear what is meant by contract.
(mm) when we talk about composition, do we always infer a transaction?
(mm) this is alluded to in the front section
(mc) transaction really means a looser sense, probably not 2PC
(mc) more of a loose "business transaction" sense
(mm) composition: is it a composition of services, of choreographies, of ...?
(mm) is there only one interface to the composed entity?
(mm) tony may want input here because it may be that composition != transaction
(mc) we should probably dumb-down the use of the word transaction
(srt) this relates to 2.1 in the first paragraph
(srt) it's not clear what is meant by transactions there
(srt) I meant services coming together to achieve a business goal, NOT a computational transaction
(mm) there's a tension here between non-business contracts, and "business" transactions
(mm) this has to do with bounding scope; when you put these in, you open yourself to further reqs
(mc) yes, we need to find some more neutral terminology
(srt) I'll take an action to recraft 2.1 to find some more neutral terminology
NEW ACTION: SRT to clarify use of “business
transaction” using more neutral terminology if possible in
section 2.1
(tony) I agree -- we should be clear with our word choice, possibly using different words
(srt) any other points or comments?
(mm) talking about generation of test messages from a chor definition choreographies deal with very dynamic behavior; could they generate executable tests?
(mc) this is an example of a usage: not a hard requirement, just potential uses
(mm) we might want to back off on what we describe as being generated here
(srt) yes, this wouldn't be part of what we deliver
(srt) BUT, it should be possible to take the definition generate test messages for services
(srt) this would be a third party tool -- Enigmatek already does this
(mm) the question is how much scope are we opening ourselves to?
(srt) we've tried to be careful to use the word MAY here
(mc) Monica's saying you might not be able to generate everything
(srt) right, incomplete coverage
(tony) if you have a specification with external interfaces, I can understand doing this
(tony) isn't it slightly different with a choreography, which is a closed world?
(tony) any message you generate has to be part of the choreography
(mc) any tester would have to pretend to be playing one of the roles in the choreography
(tony) exactly -- "test messages" don't exist, all you can do is exercise the choreography
(mc) this is no different than testing anything else
(mc) I can write a test client to execute and analyze any web service
(srt) you can have a web service coded in WS-BPEL that only partially participates in a chor
(srt) all you can do with the chor is test its participation in the chor
(srt) even in that case, you may not be able to fully test dynamic behavior of that service
(mc) we could reword to say "used to aid the testing of ..."
(mm) Steve, what you just said could be useful
(srt) any other comments?
(mm) do you see (technical contract == choreography)?
(srt) difficult to answer ...
(mm) don't think we ever resolved this
(srt) not too keen on the term "technical contract"; prefer "behavioral contract"
(mc) but what if one of those participants is not a web service?
(srt) tricky, because what exactly is a web service?
(mc) WSA says a web service has to have a WSDL interface, but is this normative?
(db) it might be useful to have a chor with no web services
(db) for instance, a workflow
(db) if you don't take this into account, you can only use chor in a fully automated system
(srt) let's log this as an issue
(mc) if we say choreography is only for web services, we'd better know what web services are ...
(mch) WSA will presumably go on to become a note (non-normative)
(srt) any other comments?
(mm) is tomorrow to late for email comments?
(srt) no, please keep the comments coming. they don't go into /dev/null ...
(srt) we're trying hard to keep up with them.
(srt) the requirements document is a living document; it doesn't freeze until we're done
(mc) but the intention is that we can sign off on something as a (relatively) acceptable version
(tony) before leaving the requirements document, how about the process going forward ...
(tony) the front-end looks good, but the back-end incldues some (potentially) major sections empty
(tony) are you planning by the F2F to have written text for sections that don't exist yet?
(mc) the main TBD is the cross-referencing of requirements to use cases
(srt) at the Sept. F2F there was a lot of discussion of classification
(srt) some entries are marked as "reclassify" with a suggestion
(srt) others just say "reclassify", with no clue as to how
(mc) example: we have no requirements in the "security" category
(mc) so the question is whether we need any, or if we can avoid this category
(srt) when we come back with the next rev. of edits, we won't take out categories willy-nilly
(tony) one thing would help is to know what the intent is on these categories (e.g. system)
(tony) some categories are very obvious (e.g. security), but not all are
(mc) the only reference we has is the WSA reference document
(mc) Daniel's been largely responsible for this, and he's not been around
(mc) if that doesn't help, send an email
(mch) advice: to avoid the situation that WSA is in now, be draconian about paring down
(mch) people vote with their energy for what's in the document
(srt) +1, if people do want these sections to remain in, they must step up
(mch) 0-based requirements: the bare minimum is the starting point
(srt) any final comments?
(ivana) stepping back to use cases for a moment ...
(ivana) regarding use case 1, variation 3 (out of sequence messaging)
(ivana) in asynchronous communications, it's possible to receive out-of-order messaging
(ivana) example: email out of sequence
(srt) if there's a contract to behavior, and someone gets it wrong, it's wrong
(mc) your choreography has to model valid sequences of communication
(mc) however you've modeled it, you need to be able to enforce that
(mc) "engine" should be added on to the end of the requirement
(mm) it needs to be made clearer: how and where this functionality is provided is not our job
(ivana) this is a question of a messaging layer, and an execution engine
(mc) the requirement needs to use the word "may"
(ivana) but then it's no longer the responsibility of the choreography language
(mc) we need to move on in the call, send comments over email and we'll discuss
Issue tracking (progress/review)
Nothing further to report
BurdettML
OracleML
(db) planning to get together early next week to go over this
(db) then we'll plan to get everyone up to speed on what's going on
(mc) looking forward to the editors producing the first version
(srt) anything to report on liaison activities?
NEW ACTION: chairs write a report for WS-BPEL technical committee prior to our F2F
(mc) it would also be nice if we could invite them to provide reports for us, as well
(srt) (another action item: chairs to request reports from WS-BPEL chairs)
(srt) Howard Smith's paper: do we want to provide comment to this? Any comments from people on this?
(mc) It's nice to see external people paying attention to our list. it's also nice to see the quality of the responses to it. Probably don't need to respond as a group ...
(mm) Is there maybe some information there that could be valuable to us?
(mm) question: how far is Cambridge from Heathrow? can you provide some info?
(srt) I've sent a map to Yves to post to the registration page
(srt) Cambridge is on average a 90-minute drive from Heathrow, no tube
(srt) need Heathrow Express Train to Paddington (takes 15 mins)
(srt) take taxi or tube to King's Cross (takes who knows)
(srt) then take the train direct to Cambridge (takes 45 mins)
NEW ACTION: SRT to edit section 2.1-2.3 to make clear what is meant by contract.
NEW ACTION: SRT to clarify use of “business transaction” using more neutral terminology if possible in section 2.1
NEW ACTION: chairs write a report for WS-BPEL technical committee prior to our F2
ACTION: PLEASE Let SRT know if they are planning to attend the F2F as they need to know the accommodation in the college.
SRT: venue IS booked, registration page should be up soon.
(IN PROGRESS, 1111, 1118)
ACTION: Steve/Yves – set up a registration page for the December FTF (IN PROGRESS, 1111, 1118)
ACTION : chairs look at WSA issues process and recommend whether it should be adopted by this group. (IN PROGRESS, 1021, 1028, 1111, 1118)
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. To be taken into account at editors meeting in November. ( IN PROGRESS, 1021, 1028,1111, 1118)
ACTION : SRT Brought semantics question to the TAG. On chairs coordination call, he asked about semantics for/of choreography. A new SemanticWeb Services Interest Group is being formed in about one month. Issue will be sent to that group when it is formed. (NO PROGRESS, 1021, 1028,1111, 1118)
ACTION: Steve RT will send a one-page summary of his thoughts ( NO PROGRESS, 1021, 1028, 1111, 1118)
NOTE THESE ARE THE SAME ACTION ITEM
ACTION: Daniel to look though document and see which requirements we captured so far regarding transactions. (NO PROGRESS, 1021, 1028, 1111, 1118)
ACTION: (Tony) Explain (in email) the substance of the issue (on transactions).
Status: (IN PROGRESS, 1028, 1111, 1118)