Notes from the OASIS WSRF TC Face-to-Face
meeting
27th June 2005
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=7739
NB: The
meeting is not quorate, not having the necessary 50% of voting members.
Confirm minute taker
Tim Banks is taking the minutes.
Approve of minutes of June 13th Telecon
See: http://www.oasis-open.org/committees/download.php/13216
There were comments and no objections.
Call
for AOB
(IanR) Conversion of members to voting
members has taken place. Conversion of observers to non-voting members to allow
contributions to be made can be achieved via a note to one of the co-chairs.
Action Review
(Sam) To post PR version Basefaults. Done
(Steve G) Raise new issue for further relaxation of RPDoc constraints. Carried fwd
from 6th June. Carry fwd.
(IanR) I suggest removing this as an
action, and awaiting a real issue.
No Objections.
(Tim) Post description of new issue to the list, describing gap in the
framework with potential resolution in SG. Done. (Issue 123).
(Bryan) Raise this new issue as
‘open’ Done.
(TomM) Raise new issue to describe the equivalent of cut/copy/paste for
metadata. Done. (Issue 118)
(TomM) Raise new issue to discuss the name of the metadata container and its
extensibility capability beyond RP documents. Done (Issue 119)
(Bryan) Raise this issue as ‘Open’.
Done
(TomM/Bryan) Write issue proposal for metadata descriptors to refer to GED and
have open attribute content. (& Move to open) (See below)
(TomM) Post revised draft of specification. Carry Fwd.
New
issues to consider - Bryan
WS-A URL:
http://www.oasis-open.org/apps/org/workgroup/wsrf/email/archives/200506/msg00043.html
reopen 110 -
http://www.oasis-open.org/apps/org/workgroup/wsrf/email/archives/200506/msg00051.html
(Bryan) Issue 117 will be replaced by issue 123 from recent email: http://www.oasis-open.org/apps/org/workgroup/wsrf/email/archives/200506/msg00058.html
Action: (Bryan) Remove issue 117
Issue 118 (cut/copy/paste for metadata.):
from last week’s minutes (corresponds to email http://www.oasis-open.org/apps/org/workgroup/wsrf/email/archives/200506/msg00061.html
)
(IanR) It should be Appnotes that is
affected.
(Bryan) Right
(IanR) We need to move this from proposed to
open when we are quorate.
(Bryan) New issue 119 is about the name of metadata container and
extensibility. It is open per last week’s minutes.
[New attendees means the meeting is now quorate.]
(Bryan) Can we open issue 118 now?
(IanR) Are there any objections to issue
118 being opened?
No Objections.
Action: (Bryan) Move issue 118 to open.
(Bryan) Issue 120 is confusing. (global elements decls and open attribute
content). Is it about open attribute content?
(TomM) I didn’t recognize it from the
minutes.
(TimB) There were to be three new issues,
perhaps last weeks minutes didn’t describe them correctly.
(TomM) The issue is about the ability to
path into sub-portions of Resource Properties vs restricting to qnames to
simplify the RMD specification.
(Bryan) Is this what I created as issue 120 (Describe rules for overlapping
descriptors in metadata)?
(TomM) Yes, that’s it.
(TimB) So I will update last week’s minutes
to reflect this issue title.
(Bryan) This issue is open as per last week’s minutes.
(Bryan) Issue 121 is about RPDoc constraints, and we’ve agreed to remove
that pending a new issue from Steve.
Action: (Bryan) Remove issue 121.
(BryanM) Issue 122, says the reference to
WS Addressing was incorrect.
(TomM) I thought I allayed the concerns
that the uri will be the stable reference for the V1 namespace.
(IanR) I would be surprised if the uri
remained the same through a V1 to V2 revision.
(TomM) That was the case (it did change) for
Xpath.
(IanR) If we’re wrong in our understanding
of WS-Addressing versioning, then we’ll have to change, but right now this
isn’t a problem. I propose we should close with no further action.
(TomM) Seconded. I will forward Hiro’s
response…actually it’s at 21 June at 6:27 he sent a response, but the email address got mixed up. Hiro says
everything is ok if the w3C naming convention is as Tom expects.
(IanR) Please can the response be forwarded
to the list?
(TomM) Ok.
No objections.
Action: (Bryan) Move to closed, no action.
(Bryan) Issue 123 is about collections, and the minutes from last meeting
say it should be open.
(IanR) Right, Any objections?
(TomM) I think the proposal is orthogonal
to the issue.
(TimB) The issue we agreed was about the
gap between ResourceProperties and a Members of a ServiceGroup. We also
discussed one way to fix it which is described in the issue. We should open it
and discuss resolution when we review the issue.
(TomM) I think the removal of a memberService
is a bit tortured, but the issue is worthy of discussion.
(IanR) Any objections to moving this to
open?
No objections.
Action: (Bryan) Move to open.
(Bryan) Also there is a discussion about reopening issue 110.
(IanR) Issue 110 is about adding open
content to the base fault type, which means that schema validation breaks. Are
there any objections to reopening this?
No Objections.
Action: (Bryan) Reopen Issue 110.
Issue review – Chair
WSRF110 Base Faults extensibility
(IanR) What are the three options for
solving the original problem?
(TomM) There were originally three options
for resolving this issue.
- Remove the extensibility from base
faults- closing issue 100 with no action
- Remove the extensibility element prior to
a required element to make it deterministic.
- Wrapper the extensibility within another
element.
(IanR) Which of these supports the original
requirement?
(Bryan) One requirement is a generic fault handler, another is the desire
to add additional elements. A generic Fault handler is possible via xml
‘extends’ without adding additional elements such as stack trace via
extensibility that aren’t defined within the fault type. (IanR) If one uses
xml ‘extends’, then surely the baseFault type can be extended, and we become
non-deterministic wrt the UPA constraint.
(TomM) If the any extension point went
above the timestamp, then the ‘any’ would be distinguishable.
(IanR) Doesn’t ‘extends’ require new
elements to be added at the end.
(TomM) Yes.
(IanR) So as long as any isn’t at the end
the ambiguity is removed.
(TomM) Yes.
(IanR) ..but there would be no access to
the elements beyond ‘any’ by a generic handler nor cope with xml ‘extends’ if
those elements go at the end.
(TomM) If processing=lax, then you don’t
have access to extended elements through the ‘any’. Right?
(IanR) How would I write a generic handler?
(BryanM) The handler might not know the
exact detailed type, but it could still generically process the basic fields.
(BryanM) We need to talk about the two
things separately.
(TomM) Yes: generic fault handling and basefault
extensions.
(IanR) of the three proposals: 1) Doesn’t
give extensibility of the core basefault type.
What is the shortcoming of moving the ‘any’
earlier? Or of Wrappingit with an element that has minoccurs=0.
(TomM) I think they both achieve the same
thing: it only deals with elements that are deterministically added. The extra
stuff can’t be accessed.
(IanR) So neither option 2 or 3 solve the
problem. So we can pick the one we like better, or drop the modification to baseFault
(Bryan) I didn’t want to insist that we had access to the extensions.
What I want is that a handler can handle generic faults. Option 2 or 3 is
acceptable, but 2 is aesthetically better.
(TomM) Agreed.
(IanR) Are there any objections to moving
the ‘any’ to be the first element in the baseFault (since there is only one
required element).
(BryanM) Other specs put extensibility
point first for this same reason.
(IanR) So the proposal is option 2. Any
objections?
No Objections.
Action: (BryanM)
Move issue 110 to resolved.
WSRF118: How do you cut/copy/paste portions of metadata
(IanR) We should discuss this on the basis
of a proposal for new text.
Action: (TomM) Write new text.
WSRF119: What should we name the metadata container, and
does it need extensibility?
(TomM) If we call this PropoertyMetadataescriptor with extensibility, how do we
circumscribe the possibilities for extension.
(IanR) We would be giving them a way to do
property metadata, but is this insufficient. Ie, what do we lose by not having
the extensibility, and how difficult is it to explain if it’s there?
(TomM) We could put extensibility into the
property element instead of in the container, so the extensibility has to be
about the property rather than other things.
(IanR) That sounds good. Anyone else have
feelings about it.
(Bryan) Since we are only talking about metadata for properties, it seems
the container should only allow property metadata.
(TomM) Right. That sounds like agreement.
(IanR) What’s the resolution?
(TomM) The action would be to rename metadataDescriptor
to resourcePropertyMetadataDescriptor, and remove extensibility from it and add
extensibility to the property element which is a child sequence of RPMD.
(IanR) Any Objections.
None.
Action: (Bryan) move to resolved.
WSRF120: Describe rules for overlapping descriptors in
metadata.
(TomM) This is straightforward if the rules
are about qname references. Ie drop pathing, use qname references and open
attribute content. This makes things much simpler.
(IanR) This is an issue because we hadn’t
captured this in the review of RMD, yes?
(TomM) Yes. In the non-quorum discussions,
we decided it would be simpler to drop pathing portion of property metadata. So
we needed to bring it to the TC.
(IanR) I think we talked about it on the
last call (or the one before). Does anyone think we need more discussion?
No-One.
(IanR) The proposed resolution is to drop
the pathing and have open attribute content which can be used to do
extensibility. Any Objections?
None.
Action: (Bryan) Move issue 120 to resolved.
Straggler
Roll Call & Close
Closed 13:04 Eastern time
Next telecon is on July 11th.
Summary of actions
(TomM) Post revised draft of specification.
Carried Fwd from June 27th.
(Bryan) Remove issue 117
(replaced by 123).
(Bryan) Move issue 118 to open.
(Bryan) Remove issue 121.
(Bryan) Move issue 122 to closed, no action.
(Bryan) Move issue 123 to open.
(Bryan) Reopen issue 110.
(Bryan) Move issue 110 to resolved.
(TomM) Write new text for resolution to
issue wsrf 118. (cut/copy/paste for RMD to go in AppNotes.)
(Bryan) Move issue 119 to resolved.
(Bryan) Move issue 120 to resolved.