Notes from the OASIS
WSRF TC
Teleconference
18th Sept 2006
Agenda
See:
http://www.oasis-open.org/apps/org/workgroup/wsrf/event.php?event_id=11825
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=11825
The
meeting was quorate.
Confirm minute
taker
Tim Banks is taking the
minutes.
Approve minutes of Teleconference on
26th June
See:
http://www.oasis-open.org/committees/download.php/19036
There were no comments
on the minutes and no objections to approving them.
Call for AOB
None.
New Issues
(Ian) there is one new
issue from the RMD PR: Define well-known MEX dialect URI name for
resource metadata descriptor.
(DaveS)
Strictly-speaking this is not on the public list. It can easily be
made public, but it changes the options slightly on how we deal with
it.
(Ian) I propose we
should open this issue. Any objections?
None.
(Ian) In Brian's
absence, I will update the issues list.
Action(Ian):
Move to Open.
Issue resolution
RMD
PR: Define well-known MEX dialect URI name for resource metadata
descriptor
Most recent discussion:
http://www.oasisopen.org/apps/org/workgroup/wsrf/email/archives/200608/msg00009.html
(Dave) This seems
logical, but we don't have refs in our specs to other specs that
haven't gone through standards bodies. I don't want to reopen the
argument by introducing a ref.
(TomR) Why can't WS-Mex
define the URI? It does so for WSDL and schema.
(Dave) Otherwise a
joint profile/pattern for Mex and RMD would specify the URI.
(Ian) The MEX authors'
perspective might be that they can't specify all uses of the spec.
(Dave) Right, which is
why a pattern might the right place.
(Ian) So, from the
perspective of this TC, how do we resolve the issue?
(Dave) We document that
it is out of scope, and that the referenced spec is outside of the
'normative' reference. Also, we aren't (so far) obliged to resolve it
(since there is no public comment).
(Kirk) So, at the
moment, I can use whatever dialect URI I like?
(Ian) And this would be
an interoperability problem?
(Kirk) Yes. The
WSDM-CIM mapping requires integration of MEX and RMD.
(Dave) So this would be
a good place to document the dialect URI.
(TomR) Yes, that would
have some industry clout.
(Ian) So the proposal
is to resolve the issue with no action, but to observe that the
DMTF's WSDM- CIM mapping can specify the URI.
(Dave) Yes,. I propose
that.
(Kirk) And I second it
(Ian) Any objections?
Action (Ian)
Close with no action.
[Lilly Liu and John
Fuller joined the call].
Discussion of WS-RT
(Ian) There are foils on the web site
here:http://www.oasis-open.org/apps/org/workgroup/wsrf/download.php/20325/WS-RT%20and%20WSRF.ppt
(Kirk)
WS Transfer is being re-written to accommodate extensions to the WS
Transfer Operations, but supplying the MustUnderstand header and
using dialects and new child elements. WS-RT
introduces a new Get operation, and this goes a little beyond the
extensibility mechanism. Will a portType have to support both - a
kind of overloaded Get?
(Ian)
The text
is trying to accommodate today's clients using WS-RT with no body.
The op returns the whole resource representation. Also, there are
endpoints that accommodate the dialects.
We propose using the same action URI so that a client can send a Get
message to either a new or old old resource and always get at least
the behaviour of the old resource (per WS-Transfer, ignoring the
fragment expression in the body). With the new spec, there are 2 ways
to retrieve the entire resource representation - using the old
WS-Transfer with no body or using the new version with a wsrt:Get
body containing no wsrt:Expression.
(DaveS)
It sounds miraculous.
(Kirk)
I find the business
of the two Gets curious.
(Ian)
Yes: Curious and Miraculous. We know this will improve when the spec
goes to a standards body and we don't have to worry about
interoperability with old clients.
(DaveS)
The Servicegroup,
BaseFaults and RMD parts of WS-RF aren't covered in the WS-RT
roadmap. Are there any plans to address this?
(IanR)
We haven't managed to agree on Basefaults, and ServiceGroups haven't
come up on the agenda.
Straggler Roll Call
See Meeting record.
AOB
(Ian) Next telecon is
potentially in two weeks, but depends on new RMD comments. We will
cancel if there are none.
(Dave) When does the
RMD PR end?
(Ian) There is at least
another month.
Closed 5:45.
Summary
of actions
(Ian) Open new Issue
to describe
'Define well-known MEX dialect URI' and move to Closed, no action.