See also: IRC
log
Present:
Steve Graham GGF
Philippe Le Hégaret W3C
Jonathan Marsh Chair (Microsoft) (x:35)
William Vambenepe Hewlett-Packard
Regrets:
Umit Yalcinalp Oracle
Chair: Steve Graham
Scribe: Philippe Le Hegaret
next week:
... OGSI ServiceData review
following week:
... WebDAV review
Ideas
for the proposal:
... 3 parts (initial
proposal, later amended)
... - Conceptual overview of the approach (why are we doing this)
... - how would you model a state in a service interface, implications on
inheritance and bindings.
... interface that models the operation to get the
data, and query it
... - interface that models the operation to get the data, and query it
William: differences between the
last 2?
Steve: first is how to model (XSI type, ...). how to define an
interface to deal with the state.
William: first part is ok. second part: abstract model of the states. third is syntax changes.
... changes to define the attribute, and let the consumer
how to access it.
Steve: agree. second
part should be a combination between abstract model and implication.
Steve: 3rd part: abstract model for
getters/access + syntax
William: would prefer to see the
abstract model in part 2, and define the syntax in 3.
... use cases and requirements?
William: use cases are not always
useful but requirements are clearly necessary.
Steve: requirements in section 1,
then?
William: maybe as a separate
document
Philippe: would still be good to
have use cases to present the idea to the group.
Steve: I can harvest those from the
OGSI group.
William: I can get some requirements
from the management work as well
William: do we have one more
specific document from OGSI or a "latest" work?
Steve: the latest version is in
public comments
... http://prdownloads.sourceforge.net/ggf/draft-ggf-ogsi-gridservice-29_2003-04-05.pdf?download
Steve: I don't expect changes to it.
GGF is meeting at the end of June. I don't expect much discussion on that
before July
William: is there one section that
contains everything?
Steve: you might need to read about
OGSI serviceData
Question
regarding including notification into our requirements
Steve: leaning towards not including
notification for the moment
Philippe: is query fundamental?
Steve: yes, not having them wouldn't
make attributes powerful.
Philippe: you can stil make them as extensions.
... we can make attributes part of the WSDL Core, and
have query as extensions.
Steve: 3
work items:
... - requirements statement, with or without use cases.
... - actual proposal (3 sections)
... - actual concrete proposal: different packaging options proposed to the WG.
Jonathan: important to present how
the TF came up with those proposals.
... and list why other solutions were eliminated
Jonathan: concerns about the need
for serviceData need to be addressed as well
<Philippe>
http://lists.w3.org/Archives/Public/public-ws-desc-state
... (mailing list will be
used for the documents for the moment)
Jonathan:
from the CG, IPR concerns mainly.
... is there a patent policy in the OGSI?
... if the attribute stuuf
goes in the WSDL, it needs to RF (following the CPP).
Steve: i.e. no legal bind if adopted
in the WSDL WG?
Jonathan: yes, no essential concept
covered.
... CPP: http://www.w3.org/TR/patent-practice
... ACTION: Steve to check IP restrictions with OGSI
ACTION: Steve to check IP restrictions with OGSI
ACTION: Steve to find a
CORBA IDL expert
ACTION: Steve to get
requirements from OGSI
ACTION: William to get
requirements from the management prospective