Prelim Minutes of OASIS WS-RX Teleconference

March 18, 2006

Start Time:4:00 PM Eastern Time

Paul Freemantle acted as chair.

Textual Conventions

� Action Item

Motion

� Resolution

1 Roll Call

From Kavi:

xx of 49 voting members, meeting quorate

Tom Rutt agreed to take minutes.

2 Agenda Approval

Date:� Thursday, 16 March 2006

Time:� 01:00pm - 02:30pm PT

Event Description:

Dial-in:

TBD

IRC/Q Mgmt(thanks to DougD): http://webconf.soaphub.org/conf/room/wsrx

Agenda:

1) Roll Call

2) Review and approval of the agenda

3) Approval of the Mar 9, 2006 meeting minutes

�� http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.php/17132/MinutesWSRX-030906.html

4) AI Review

�� http://www.oasis-open.org/apps/org/workgroup/ws-rx/members/action_items.php

5) Raleigh F2F Planning

�a> Agenda bashing

6) New issues since last conf-call

�� Watch for Marcs email

7) Issue Discussion:

� a> i021�� An RM Policy applies two-way

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i021

� b> i008�� Policy assertions granularity

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i008

� c> i090�� Use of offered sequences unclear in current spec

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i090

� d> i089 suggest the restricted use of anonymous URI

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i089

� e> i096 Complete the state tables

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i096

8) Any other business

Jeff M asked that issue 21 be postponed to the f2f..

Paul C suggested we discuss 21 and then put the best proposal on the f2f agenda for final descision.

Marc G: I agree, the concensus is on a modified version rather than the current proposal.

Agreed to discuss but not make final decision, with on objection from Jeff M.

3 Approval of the 3/9 minutes;

�� http://www.oasis-open.org/apps/org/workgroup/ws-rx/download.php/17132/MinutesWSRX-030906.html

Chris F stated that his corrections were not incorporated.

Will approve at F2F.

4 AI Review

�� http://www.oasis-open.org/apps/org/workgroup/ws-rx/members/action_items.php

action item 56 left open.

5 Raleigh F2F Planning

�a> Agenda bashing

Doug D asked if anyone not in ballot is attending;.

Sanjay asked that this be taken off-line.

Sanjay asked for comment on http://www.oasis-open.org/apps/org/workgroup/ws-rx/email/archives/200603/msg00099.html

Umit asked for a tweek on the schedule for critical issues.

Paul F stated that state tables should not be discussed before outstanding issues.

Tom R asked for a discussion on an application notes document, Fujitsu will provide a contribution before meeting.� It was agreed to put this on the end of the agenda.

6 New issues since last conf-call

�� Watch for Marcs email

No comments received.

7 Issue Discussion:

7.1 a> i021�� An RM Policy applies two-way

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i021

Chris F had a most recent posting at: http://www.oasis-open.org/apps/org/workgroup/ws-rx/email/archives/200603/msg00068.html

Which was against Sanjay proposal at: http://www.oasis-open.org/apps/org/workgroup/ws-rx/email/archives/200602/msg00222.html

�

Sanjay asked for clarifications on whether optional means the sender must be an RMD as well as RMS.

Chris F: this applies in both directions, it could be off or on in either direction, given the optional label.

Sanjay: does this impose constraints on client.

Chris F: I think the optional assertion means each endpoint gets to choose.

Marc G: I am not sure this is the latest proposal.� Paul had a more recent proposal. �What is the difference of his proposal from what is on tab.e

Paul F: I am happy with the proposed interpretation of wsrm optional. ��Anish�s concern is that if optional is place on return, the client would have to support both alternatives.

Chris F: I disagree because the client made no such assertion.

Chris F: If client wishes to expose wsdl, it could choose to not decorate with assertions.

Paul F: I believe my proposal is the same as Chris F with exception of the interpretation of wsrm required optional.

Marc G: I prefer Chris F interpretation of wsrm required optional.

Jacques D: I want to ensure the semantics of optional include the ability to deal with sequences. �If the server is unable to open a sequence to me (failure in first instance), if I send message in non reliable will the message be accepted by server.

Chris F: If I am server, and decorate wsdl with endpoint polkcy wsrm assertion = optionale sender can use it or not.

Jacques: the outut messages have wsdl decorated as optional on output, the server decides to send within a sequence and the client is unable to do that.

Chris F: if the server cannot establish a sequence it should fall back to unreliable exchange.

Jacques; given Anish concerns the optional qualifier means that the agreement can go either way. �This should not ever require the use of rm to be required. �If the client and server decide that a reliable exchange over policy that is optional, and want to indicate that the client should not fallback to non reliable qos.

Sanjay: are not such contracts out of scope.

Jacques: but if optional means either reliable or not reliable, means such an agreement that I cite is not available.

Umit: I would agree with Sanjay. �If out of band policy is stricter than what is expressed it will cause extra problems.

Umit: I ask question to Chris F on details.

Chris F: I �want a way to state that reliability is required on every message.

Umit: I am okay with the wording of Chris F>

Chris F: in response to Jacques;

when choosing between optional policy statements, there is an implicit merge on the part of the client between the policy exposed by the service and its own, internal policy

Chris F: with optional, if either side wants to use it, or it is not available, it should fallback to non reliable.

Jacques: if the client does not support reliability in response, the server should fallback on unreliable return of message. �Is there a case to not fallback to unreliable?

Sanjay: the TC seems to be reaching concensus on Chris F proposal. �We should stop discussion today.

Marc G: are there any changes needed based on this discussion to the proposal from Chris F. �The agenda should have links to the proper proposals.

� Action: Chris F will send a consolidated proposal by Monday evening, incorporating any comments sent on the list.

7.1 b> i008�� Policy assertions granularity

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i008

Doug D: can this be closed.

Tom R: the current solution to 21 has message level granularity, so is this allowed.

Marc G: I would say it is .

Paul C: we should put both of these on the agenda on f2f.

Agreed to put issue i008 on F2F agenda.

7.2 c> i090�� Use of offered sequences unclear in current spec

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i090

Matt proposal: http://www.oasis-open.org/apps/org/workgroup/ws-rx/email/archives/200603/msg00132.html

Paul F: I posted a link. http://www.oasis-open.org/apps/org/workgroup/ws-rx/email/archives/200603/msg00132.html

Paul F: I want to remove text that offer is just an optimization. �I would like to deal with this outside of anonymous case. �Anonymous with offer should be a separate issue dealing with linking �of lifecycles of both sequences.

Matt : we may want to link sequences, but this would be by outcome of I 089.

Paul F: the issues of offer, without anonymous case, should be what 90 is restricted to. Add new issue for anonymous case.

Doug B: I suggest we wait �for the resolution of I089efore posting a new issue.

Doug D need a nit change:

MattLovett Change to

Line 282 Insert

/wsrm:CreateSequence/wsrmffer/wsrm:Endpoint

This REQUIRED element, of type wsa:EndpointReferenceType as specified by WS-Addressing [WSAddressing] specifies the endpoint reference to which WS-RM protocol messages related to the offered Sequence are to be sent.

Marc G: with regards to a new issue regarding offer with anon, I agree this is a separable concern.� The lifecycle linkage of the two sequences in the offer case could be a separable issue.

Matt: I agree to remove the optimization text, but it does not have to be done under this issue.

Marc G: if we can bundle that into this proposal I can support it.

Matt: I believe accepting a basid proposal here can simplify the resolution of I089.

Chris F: I would like to move this, however we need to have the word changes from Marc G be accepted.

Matt posted:

MattLovett Addition to the proposal: Line 238 strike: Note that

offering a Sequence within the <wsrm:CreateSequence> element is simply a protocol optimization.

Marc G: from lines 236 thru 241.

Chris F moved to close I 90 with amended proposal, Marc G seconded.

� No opposition, issue 90 closed with amended proposal.

7.3 d> i089 suggest the restricted use of anonymous URI

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i089

Doug D asked that we defer discussion of I089 to the Face To Face meeting.

Paul F: I agree that based on mailing list discussions, there have been circles. �At the f2f we should have the issues presented that the TC needs to resolve.

Paul C: will this be covered under issues list or interop.

Paul F: it should be covered on the issues list.

Paul F� I suggest Doug D and I give a presentation on this issue.

Paul C: could this discussion occur when the interop people are still present.

Paul F: we will try to work with that comment during the agenda finalization.� Some of this should be discussed while interop SC is present.

Tom R: Is it agreed that the recent discussions from Interop SC on anonymous with synch response is part of the subject of Issue 089.

Paul F: I agree that that is part of the resolution to I089.

Doug D: I do not want to ban anonymous reply to across the board. �I need to float my ideas across the board before a decision is made.

Umit: The important thing is to decide on a solution first, then to address whether anonymous meets that solution.

Marc G: a presentation with message flows would help us come to a resolution.

Tom R: some sequence timing diagrams would be usefule

Paul C: it would be good to have someone on point for all outstanding issues for the F2F. �It would be good to have a presentation on I021 as well.

Sanjay we have two issues 89 and 21.�

Tom R: a set of message sequence diagrams would help.

Paul F: I intend to present such diagrams.

7.4 e> i096 Complete the state tables

�� http://docs.oasis-open.org/ws-rx/issues/ReliableMessagingIssues.xml#i096

Paul C: do we have someone to fix the existing problems.� The problems I posted are simple and generic. �If we do not have the work done by f2f we would have to recess.

Doug D: I will work to get this done by the face to face.� I will accept being owner on this one.

Paul C: there are two items, agreement with cd 3,and answer of the pro-forma concerns that have arisen.

8 Any other business

Chris F: do we have plan of action out of Face to Face.

Sanjay: we have time to discuss this on the agenda.

Paul F: we have allocated an hour on the F2F.

Paul F: at the last meeting we allocated two weeks for a next CD with a ballot on that with public review as soon as the ballot passes.

Paul C: can we have the adjusted proposal in an email for discussion before the f2f.

Paul F: I can do that.

� Action: chairs will send email on adjusted schedule for progression before the f2f.

ws-rx message