Notes from the OASIS
WSRF TC Teleconference on
23rd August 2004
Roll call
The roll call is kept on the TC web site
under the meeting record.
See http://www.oasis-open.org/apps/org/workgroup/wsrf/event.php?event_id=4801
Approval of minutes from previous telecon (9th
August)
See: http://www.oasis-open.org/committees/download.php/8656
There were no comments and no objections to approving the
minutes.
Other Action Review
Action Review.
(SteveG) Propose text to resolve issue 4. (Carried fwd from f2f)
Action: Defer to next call
(TomM) Write resolution for issue 10. (Carried fwd from f2f)
Action: Defer to next call
(Sam) Raise a new issue to consider an identity mechanism. (Carried fwd from
f2f)
Action: Defer to next call
(Bryan) Move issues 62 through to 64 to Open. Done
(Chairs) Add agenda item for next face-to-face: discuss the scenario doc and the
schedule. Done
(Ian) Add Agenda item for the next call to review Requirements doc Done
(Ian) Add Agenda item for the next call to review outline for the App Notes doc
Done
(SteveG) Make a proposal for text to resolve issue 1 (Origins of interface
aggregation) to mailing list. See discussion below
(Bryan) Move issue 1 to resolved.
(Ian) Should we move back to open
until precise text is available?
(Bryan) Yes.
Action: (Bryan) Move issue
1 to ‘Open’
(Glen) Propose wording to resolve issue 20 (Notification message format) to
mailing list.
Action: Defer to next call
(WilliamV) Start a mailing list discussion for issue 24 (xpath Namepsace).
Action: Defer to next call
(Ian) Continue discussion of issue 24 at next telecom. Done
Acceptance
of New Issues to the issue list.
Issue WSRF67: TerminationTime property should
allow Set/Update
(SteveG) Is there doc in the spec that explains why this is
not allowed?
(Latha) There is text to explain it, but not why.
(DaveS) Semantics of setting the time are different in the
explicit op – eg synching of the clock.
(Latha) It is strange that this property setting needs to be
disabled.
(DaveS) But, if we have metadata for set-ability, that
anomaly goes away.
(IanR) Doesn’t everyone agree it’s not an issue, because we
don’t need two ways to set termination time?
(Bryan) I raised it on behalf of implementers. They think it
looks like a strange anomaly needing extra work to implement.
Action: (Bryan) to review with implementers. Leave status
as ‘proposed’.
Issue ???: ResourcePropertyChangeNotification
See: http://www.oasis-open.org/archives/wsrf/200408/msg00043.html
Need to put property value in the notification
(No objections)
Action: Move to open
Issue ???: using setResourceProperties
See http://www.oasis-open.org/archives/wsrf/200408/msg00044.html
Contact – Roger Mendez
(DaveS) Propose email discussion via the list.
Agreed.
Action: (All) Discuss via maling list.
Requirements document
(DavidL) There should be something next week. Review at the
next call.
Action:
(Ian) Add Agenda item to review requirements doc at the next call.
Review outline for the App Notes doc
The outline was proposed in:
http://www.oasis-open.org/archives/wsrf/200408/msg00000.html
(TimB) The AppNotes outline is different in scope from the
Primer: more sophisticated. A Primer would be an introduction, including the
kind of material in the whitepapers. (AlanW) But we already have the
whitepapers.
(IanR) Face-to-face showed some interest in the Primer
concept: perhaps there is a need for two documents?
(TimB) What will happen to the Whitepapers?
(IanR) The whitepapers are needed and some material will be
brought forward into the new normative document, but we don’t have anything
like the “State” paper and/or OGSI Primer.
(Igor) It’s useful to put the updated papers into a TC
document. The application note clarifies intricacies.
(IanR) Are there volunteers to support/help review the
Primer?
(Yes: Igor/Katy/Jem/Alan)
Action: (TimB) Propose Primer outline for next call.
(AlanW)Is the AppNote outline approved?
(IanR) Proposed: Outline for the appnote is approved
(No objections)
Action: (AlanW) Continue to generate contents for AppNote.
Issues review
(SteveG) If we try to merge WSRF specs with domain-specific
interfaces we lose information about the origin of the operations.
Approaches: we do not have a composition model – can investigate
the portType – but this is hard for the client.
(DaveS) Even this requires a hope and a prayer – the
semantics may not be the same.
(DavidL) The inference gets larger as new versions of the
portType are produced.
(SteveG) Other options are a full-scale backport of WSDL
2.0, or an attribute on the portType – which is a halfway house. If we call it
‘extends’ then we might be confusing it with the attribute used in WSDL 2.0. We
can keep the portType as it is, but add documentation elements on the
operations. However, this needs tooling to look in the new elements.
(Igor) Why is this useful?
(SteveG) Because (eg) if I’m discovering servers to manage,
I can’t tell which portType is being extended, and so which semantics are
implied.
(Igor) This doesn’t seem to be the job of WSRF. Why isn’t
the responsibility of WSDL?
(Fred) We ‘object’ people think extension is a good thing, but
couldn’t this be done with multiple ports.
(SteveG) Some tooling may not cope, and addressing
mechanisms may not be compatible. Eg WS-Addressing does not take a port/service
view.
(Igor) We can equivalently rewrite wsdl 1.1 multiple ports
with different portTypes as WSDL 2.0 (?)
(IanR) Can we ask the WSDM TC (as the proposer) for
preferences about how this should work?
(Igor) WSRF is not a trash can for all the issues that WSDM
has. This is a Web Services issue, and it belongs with WSDL 2.0.
(SteveG) We can’t wait for WSDL 2.0.
Actions:
(SteveG) Write proposal
about ‘extends’ on portType
(Igor) Write proposal about (?)
(Fred) Write proposal about
multiple ports
WSRF48: Specify
behavior of nillable properties
Clarify status of this issue. This has been targetted as an AppNotes
issue and a proposed
resolution posted
(http://www.oasis-open.org/archives/wsrf/200407/msg00128.html). Is this issue
resolved?
(Bryan) This is left as open, pending production of appnote/Primer.
(IanR) This can be moved to ‘resolved’ if there are no
objections to the text.
(Agreed)
Action: Move to resolved.
WSRF27: Add operation to return entire property document
(SteveG) It would be quick to produce a proposal
(DaveS) It is possible to get the whole document with a “query
all”, so do we need it?
(SteveG) The operation supports the REST mode of operation.
Also, to retrieve a whole list of properties is easier with a single op.
(IanR) Proposed to have a GetAllProperties operation, with
the same signature as the current operation.
Action: (SteveG) Proposal text.
Summary of actions
(SteveG) Propose text to resolve issue 4. (Carried fwd from
f2f)
(TomM) Write resolution for issue 10. (Carried fwd from f2f)
(Sam) Raise a new issue to consider an identity mechanism.
(Carried fwd from f2f)
(Bryan) Move issue 1 to ‘Open’
(Glen) Propose wording to resolve issue 20 (Notification
message format) to mailing list.
(WilliamV) Start a mailing list discussion for issue 24 (xpath
Namepsace). (Carried fwd from 9th Aug)
(Bryan) Review issue 67 with raisers (implementers) in the
light of telecon discussion, and leave status as ‘proposed’.
(All) Discuss potential ‘Using SetResourceProperties’ issue
via mailing list.
(Ian) Add Agenda item to review
requirements doc at the next call.
(TimB) Propose Primer outline for review next call.
(SteveG) Write proposal for Issue 1 using ‘extends’
on portType
(Igor) Write proposal for Issue 1 using (?)
(Fred) Write proposal for Issue 1 using multiple ports.
(Bryan) Move issue 48 to Resolved.
(SteveG) Propose text for issue 27.