Notes from the OASIS
WSRF TC
Teleconference
15th May 2006
Agenda
See:
http://www.oasis-open.org/apps/org/workgroup/wsrf/event.php?event_id=11816
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=11816
The
meeting was quorate.
Confirm minute
taker
Tim Banks is taking the
minutes.
Approve minutes of Teleconference on
3rd April
See:
http://www.oasis-open.org/committees/download.php/17519
There were no comments
on the minutes and no objections to approving them.
Call for AOB
None.
Action Review -
chair
(Ian
& Jem) Review new version of the Primer. Done
(IanR)
After review, set up ballot to adopt the new draft of the primer.
Done
(Bryan)
Move issue 172 to closed. No action. Done
(IanR)
Set up e-ballots for remaining RMD issues given 6-week gap between
April 3 and May 18. Resolved one,
one still in progress, plus raised a new one. See below.
Review of ballot progress (Ian)
(IanR) We've had two ballots. Issue 175 (Initial
Values) was resolved. The second ballot was to approve the updated
Primer as a new committee draft, following review by Ian/Jem, and
this ballot is still in progress and runs until next Saturday. If
this ballot succeeds, this would finish off the Primer. The essential
difference with this Primer is that it has updated links to the OS
standards, and hyperlinks to the full xml for the examples.
New Issues
WSRF176: Simplification of
metadataDescriptorLocation type
See:
http://www.oasis-open.org/apps/org/workgroup/wsrf/email/archives/200604/msg00050.html
(IanR)
This was spun off from issue 174, and proposes simplifying the schema
for the RMD spec. Currently we have pairs of URIs for
metadataDescriptorLocation for the metadatadescriptor when, in fact,
we need only one URI. The multiples hark back to the time when we had
multiple metaDatadescriptor docs and composition thereof, but now we
only have one doc.
(DaveS)
is everyone still in touch with this? Does everyone understand this?
Or rather, is there anyone who doesn't?
No-one
(DaveS)
So let's move it to open.
Action:
(BryanM) Move to open.
(DaveS)
Are there any other new issues?
None.
Issue resolution
WSRF175: Add InitialValues section to RMD
(IanR)
This was resolved by ballot:
http://www.oasis-open.org/apps/org/workgroup/wsrf/ballot.php?id=1012
Dan posted a
resolution
to include InitialValues and there was some discussion and we changed
from the term 'Default' values to 'Initial' Values.
(DaveS) We had below
50%
voting on this
(IanR) The majority
of
those voting voted yes.
(DaveS) But do we
need
a majority of voting members? What's the quorum?
(MartinC) I don't
think
the ballot was quorate. We could treat the ballot as a meeting, and
use lack-of-vote like lack-of-attendance.
(DaveS) To keep it
simple, let's have a vote right now. The proposal introduces Initial
values that are
distinct from Static values. Any questions?
None.
(DaveS) Any
objections to making a motion?
None.
(DaveS) Any
objections to the motion?
(BryanM) I want to
abstain.
(Lily) I want to
abstain, too.
Action:
(BryanM) Move to Resolved.
WSRF176:
Simplification of
metadataDescriptorLocation type
(IanR) Currently the
metadatadescriptorLocation is a list of pairs or URIs, with the
first-in-pair a namespace names, and the second-in-pair a location of
the metadataDescriptor document which populates
the namespace. However, we can have only metadataDescriptor document,
so we need to change the type of the location attribute to be a
single URI indicating the location.
(DaveS) It's not
like
we removing anything: the first URI of the pair (the namespace name)
is implicit, because there is only one.
(IanR) Right. This
is something
that just go left out when we simplified RMD a while ago.
(DaveS) Are there any objections to accepting the
resolution as described in the proposal?
None.
Action: (BryanM) Move to resolved.
(DaveS) Are there any problems with making the
change, Dan?
(Dan) No.
WSRF174: Metadata Descriptor Instance as Resource
Property
Useful
summary:
http://www.oasis-open.org/apps/org/workgroup/wsrf/email/archives/200604/msg00030.html
(DaveS) This is
about including a copy of the MDD in the resource directly. The
latest proposal is
that we (optionall) include in the Resource Properties of a resource,
an EPR to another WS Resource which the client can use to obtain the
metadataDescriptor.
There were several
options that
were discussed about the format of the target of the EPR Currently,
Tim, Ian and I have
agreement about option (d) in the email above where the EPR points to
a Definitions element.
Some of the
ramifications
are that the easiest way to get the whole document is to use
getResourcePropertiesdocument,
but the information can be got by other means using various
references such as wsdl location, interface name and namespace
prefixes.
(IanR) What this
essentially
means is that if a client has a need to get the metadata for a
particular instance, going via
the EPR and getResourceProperty-on-MDD-element will be sufficient in
most cases. All dynamic information will be there. Any static
information (on the Definitions element) can be got in other ways.
(DaveS) Right, so
there could be multiple Property Elements, and queryRP might be
needed to pick them out. GetRP
will get the whole MDD element. I think this is a reasonable
compromise over the various alternatives. Any questions?
None
(DaveS) If we do
resolve this would we be able to implement the changes and vote for a
committee draft next time?
(Dan) So, for 30th,
I think so, Yes.
(DaveS) Well, a
little
before hand so that a couple of people could look at it.
(IanR) Well 29th
is the
next meeting, and it's memorial Day. Let's target the 12th
June.
(DaveS) I will look
at it. Anyone else?
(IanR) I will, too.
(BryanM) Me
too.
(DaveS) So are there
any objections to making a motion here?
None
(DaveS) And
are there any objections to resolving this issue as described by
option (d) in the email referenced above?
None.
Action:
(BryanM) Move to resolved.
(DaveS) So, do we
have enough information to implement this?
(Dan) Yes.
(IanR) This is the
last open issue. If we can have the new RMD draft a week before 12th,
and vote for committee draft on 12th June,
then how much further should we go with it? Will we have three
companies with implementations
necessary for the oasis spec stage?
(DaveS) I think
Committee spec would be far enough, then we can see what the interest
is – we might find it hard to get the votes out at the member
voting stage.
(Dan) Do we need
three independent implementations, or three companies working on the
same implementation?
IBM and HP are participating in Apache MUSE.
(IanR) it needs to
be three member organisations. Why don't I send out the call for
ratification right now, and we can judge from the response where we
should go.
(DaveS) We need this
for WSDM, yes?
(IanR) Yes.
(BryanM) I thought
there
was some requirement from the grid community, too.
(DaveS) Right, but
the grid community is absorbed by the convergence document.
Action:
(IanR) Send out call for ratifications.
Straggler Roll Call – see Meeting record.
Closed 17:45.
The next scheduled
meeting (29th May) is canceled.
Next telecon is in four
weeks on 12th June.
Summary of actions
(BryanM) Move issue
175 to resolved.
(BryanM) Move issue
176 via open to resolved.
(BryanM) Move issue
174 to resolved.
(Dave/Ian/Bryan)
Review updated RMD ready for CD vote.
Action:
(IanR) Send out call for successful usage of the RMD spec.