[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Raw minutes of 2007-12-06 telcon
Danny van der Rijn: Good {time of day} to you all
Mike Edwards: greetings Danny
Room information was updated by: charltonb
Dial-in:
1-888-967-2253
+1-650-607-2253
+61-2-8817-6100
+44-118-924-9000
Meeting ID: 877770
Meeting password: SCABPEL (7222735)
Web-conf: http://webconf.soaphub.org/conf/room/sca-bpel-TC
Acrobat connect room URL: http://my.adobe.acrobat.com/scabpel/
charltonb: good day!
Danny van der Rijn: charlton - are you on the call?
Mike Edwards: are you on the call Charlton ?
charltonb: joining
Please change your name from 'anonymous' using the Settings button
charltonb: here
charltonb: martin are you on the phone?
Martin C: just joined
charltonb: brill
Mike Edwards: wow! a volunteer scribe !
anish: Agenda:
http://www.oasis-open.org/apps/org/workgroup/sca-bpel/event.php?event_id=16231
anish: Date: 6 Dec 2007
Danny van der Rijn: anish: agenda bashing?
Danny van der Rijn: anish: Dieter - you have a new version on the
editors list?
Danny van der Rijn: Dieter: we do have a new version. minor editing
plus inclusion of resolved issues. no objections from editing team.
will upload to TC mailing list. Can we look at it together on the next
call, approve it as a clean copy to start the year with?
Danny van der Rijn: anish: All should review the document that will be
uploaded. Sanjay should put it on the agenda for next week.
Danny van der Rijn: anish: last week's minutes..
Danny van der Rijn: martin: move to accept last week's minutes
Martin C did you mean me...i made no motion
Danny van der Rijn: minutes approved with no objection
Dieter Koenig: BPEL-12 ("Long-Running Request-Response Operations") has
now been opened in the SCA-Assembly TC as ASSEMBLY-33 - see
http://www.osoa.org/jira/browse/ASSEMBLY-33
Martin C: i think you ment Mike E
Danny van der Rijn: martin - i realized that afterwards. I'm not
familiar enough with others' voices.
anish: Topic: AI review
anish: Topic: next F2F
Martin C is amazed that danny doesnt recognize my accent by now
Danny van der Rijn: Mike E was speaking over Anish...
Martin C i know i know, we all look alike as well
Danny van der Rijn: anish: proposal to colocate a F2F with Bindings TC
in Boston week of Mar 3
Danny van der Rijn: danny: stick to original symposium meeting in April
Danny van der Rijn: martin: no objection (it *was* you this time, Martin)
Danny van der Rijn: anish: how many days do we need?
Danny van der Rijn: martin: 2. 1.5 minimum
Martin C: we will prob be the only sca tc at the sympsium which is a
good thing becuase of overlap
Alex Yiu: BTW, Issue BPEL-12 has been updated with a Dieter's comment.
Danny van der Rijn: anish: talk with Sanjay offline
Martin C mmmmm meaty things
Dieter Koenig: Issue BPEL-16 ("Ambigous Service Resolution")
http://www.osoa.org/jira/browse/BPEL-16 (originally opened as ASSEMBLY-29)
anish: Topic: New issue
Danny van der Rijn: anish: 1 new issue, raised by Dieter, BPEL-16
anish: Scribe: Danny van der Rijn
Danny van der Rijn: Danny: This is a cross-TC issue. Is dealing with
it here the right place?
anish i see the Q, letting some conversation go on for a bit
anish: ScribeNick: Danny van der Rijn
anish: Meeting: OASIS SCA BPEL TC Teleconference
Alex Yiu: Dieter ... maybe you can find more details of Java's
createSelfReference() and share with the rest of us then we can decide
what to do in BPEL ?
Danny van der Rijn: Mike E: not clear that this is an assembly problem
to solve. Could be a problem for all the C&I TCs. Assembly allows a
service to have multiple endpoints. Component won't know how many
endpoints there are.
Danny van der Rijn: Martin: let's look at the BPEL problem and solve
it. Then if we want to make an assembly proposal, we can.
Danny van der Rijn: Najeeb: How is this different when BPEL is used
standalone?
anish would entertain a motion here
Mike Edwards: I want to move this issue to be opened
anish: Present: Present: Mark Ford, Charlton Barreto, Michael Rowley,
Mike Edwards, Dieter Koenig, Martin Chapman, Anish Karmarkar, Ashok
Malhotra, Alex Yiu, Sanjay Patil, Najeeb Andrabi, Danny van der Rijn
anish oops
anish: Present: Mark Ford, Charlton Barreto, Michael Rowley, Mike
Edwards, Dieter Koenig, Martin Chapman, Anish Karmarkar, Ashok Malhotra,
Alex Yiu, Sanjay Patil, Najeeb Andrabi, Danny van der Rijn
Danny van der Rijn: Alex: We should open the issue. If Dieter can find
out more about Java createSelfReference(), that would be good.
Partially agree with Najeeb, but BPEL assumes that one active binding used.
Danny van der Rijn: Mike E: moves that the issue be opened. Charlton
seconds
Danny van der Rijn: issue opened unanimously
Danny van der Rijn: Anish: Issue 11 discussion
Danny van der Rijn: Mark: explains issue 11
anish: topic: issue 11
anish: Michael's proposal:
http://lists.oasis-open.org/archives/sca-bpel/200711/msg00019.html
Danny van der Rijn: Mark: my issue with Michael's proposal is that it's
different for literals and expressions. Either honor or ignore.
Danny van der Rijn: Michael: Had similar concern, Michael explained to
me, but it was too long ago and I don't remember
Danny van der Rijn: anish: let's come back to this later
Danny van der Rijn: topic: issue 3
Alex Yiu: http://lists.oasis-open.org/archives/sca-bpel/200711/msg00061.html
Danny van der Rijn: Mike E: Do we agree that we can determine that there
will never be a match when SCA conversations and BPEL correlation set
doesn't match for the SCA-identified conversation?
anish: s/Mike E:/MichaelR:/
Mike Edwards:
anish: danny: simple Q -- do we have anything in BPEL C&I spec that says
something about conversatoins and bpel process instance
Michael Rowley: Support for conversational interfaces
WS-BPEL can be used to implement an SCA Component with conversational
services. See the SCA Assembly Specification [SCA-Assembly] for a
description of conversational interfaces. When an interface that has
been marked as conversational is used for a role of a partner link, no
other mechanism (such as the WS-BPEL correlation mechanism) is needed to
correlate messages on that partner link, although it is still allowed.
This means the SCA conversational interface is used as an implicit
correlation mechanism to associate all messages exchanged (in either
direction) on that partner link to a single conversation. When the EPR
of the partnerRole is initialized a new conversation MUST be used for an
operation of the conversational service.
Any process which, through static analysis, can be proved to use an
operation on a conversational interface after an endsConversation
operation has completed SHOULD be rejected. In cases where the static
analysis cannot determine that such a situation could occur, then at
runtime a sca:ConversationViolation fault would be generated when using
a conversational partner link after the conversation has ended. See the
SCA Assembly Specification [SCA-Assembly], section 1.5.3 for a
description of this fault.
It is important to point out that the WS-BPEL correlation mechanism is
not restricted to a single partner link. It can be used to associate
messages exchanged on different partner links to a particular WS-BPEL
process instance.
Danny van der Rijn: Michael: also wouldn't object to closing with no action
Danny van der Rijn: Mike: What does this do to the case where they conflict?
Danny van der Rijn: Michael: Proposal is to name a fault. If we don't
standardize, each runtime could do their own thing.
Danny van der Rijn: anish: are you moving to close with no change?
Danny van der Rijn: Mike: would be just as happy to accept proposal
Danny van der Rijn: s/Mike:/Michael:/
Danny van der Rijn: alex: agree with Michael. value of proposal is
limited, given the small use case it's addressing
Danny van der Rijn: Dieter: agree that we can't cover all cases. Some
cases are more likely to happen. Event handlers going out of scope, etc.
Danny van der Rijn: anish: calls for motion
anish: danny: reason for standardizing the fault is so that it can be
dealt with it programmatically
anish: ... the number of programmer that would use this is going to be
very small
anish: ... not very useful
Danny van der Rijn: Najeeb: Other processes can use the fault
Danny van der Rijn: Michael: number of (BPEL) developers on client side
who will choose to do something different with this fault is vanishingly
small
Danny van der Rijn: Najeeb: But they can identify cause of problem
Danny van der Rijn: Michael: going to human reader, each runtime can do
better or worse job describing
Danny van der Rijn: Najeeb: can't catch them in standardized way.
Danny van der Rijn: Michael: right, but how many people want to do so?
If it's large, we should standarize.
Danny van der Rijn: Najeeb: Would like to know what happened if using
large numbers of correlations.
Mike Edwards: +1 for a vote
Danny van der Rijn: Alex: move to accept proposal here
http://lists.oasis-open.org/archives/sca-bpel/200711/msg00061.html
Danny van der Rijn: Michael: second
Sanjay: 5 mins left
anish: /thx sanjay
Sanjay: before the call ends
anish to the chat nick
Sanjay: You MAY be SHOULD unless you MUST
MUST, MAY OR SHOULD? ?
anish if we run out of time, the motion expires
Danny van der Rijn: alex: partially agree with Danny's concern. Proposal
is to standardize a fault vocabulary. But we could define cases for
conformance test
Danny van der Rijn: Michael: could define test. could test that fault
is thrown or not
Danny van der Rijn: Alex: since we agree that it's an optional feature,
test could determine if implementation implemented optional feature
Danny van der Rijn: anish: can we approve unanimously?
Danny van der Rijn: Danny: objects
Mark Ford: we may have lost some people
MUST, MAY OR SHOULD? is lost in space
Danny van der Rijn: mark no
michael abstain
mike yes
dieter abstain
martin
anish abstain
ashok
alex yes
sanjay yes
najeeb yes
danny no
simon abstain
Danny van der Rijn: Anish: motion carries 4 - 2. Issue 3 is resolved
Danny van der Rijn: meeting adjourned
anish danny, thx for scribing
anish u get a free trip to the bottom of the scribe lit
anish or even lit
anish or even list
anish can't type
charltonb: sorry, called away on an urgent matter
charltonb: btw Alex, +1 (yes)
anish: Present+: Simon Moser
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]