OGSA F2F meeting Reference Model - 15 August 2007
=================================================

* Participants

Hiro Kishimoto
Andreas Savva
Michel Drescher
Mark Morgan
David Snelling
Donal Fellows
Steve McGough
Vesselin Novov
Fred Machiel
Chris Kantarjiev
Steven Newhouse
Ravi Subramaniam (phone)

Note Taker: Hiro Kishimoto

* Reference Model Overview

Dave goes over overview slides

Document schedule

AI: Dave to update 2nd document title by deleting word EGA and
    living charter as well.

Relationship UML diagram is somewhat similar to CMDBf and also
CIM-core. Even though CIM-core does not have a instantiation link.

Lifecycle model is in the lifetime model.

Hiro asks if Fujitsu and eBay will develop this spec, have
interop test and write an experimental document in order to
promote to full recommendation spec?

Dave answers RM is rather abstract spec and concrete rendering
is necessary for interoperability. RM-WG has discussed it is
GFD-I or GFD-R and wants to GFD-R. One possibility is having
rendering specification as an appendix.

Ravi asks if RM-WG treat Policy as black-box?
David answers it will be very much black-box.

Steve is still unconformable with RM Lifecycle diagram.
Even if BES stopped to accepting creation request,
it is still in Active state.
BES is rather going from blue to green directly and no
need to trigger yellow to green.
Steven thinks meaning of extant/commissioned/active varies
service by service.
Dave agrees. EGA has more than 200 use cases and they are all
mappable to this diagram.

RM-WG is still working on crisp definition of three states.
Steven is not comfortable especially Extant. Its old name
is Unconfigured.

It is manageable if it's in Extant state.
RM-WG has delete sub-state commitioning.

Steven thinks common RM lifecycle works for infrastructure
but not good for upper ones.
Dave says it is true since it comes from EGA but he is trying
to extend.