OGSA Teleconference - 7 March 2005
==================================

* Participants

  Dave Berry (NeSC)
  Ian Foster (ANL)
  Andrew Grimshaw (UVa)
  Masayuki Hatanaka (Fujitsu)
  Yoshiaki Hirotani (NEC)
  Soonwook Hwang (NII)
  Hiro Kishimoto (Fujitsu)
  Tom Maguire (IBM)
  Mark Morgan (UVa)
  Takuya Mori (NEC / ANL) 
  Steven Newhouse (OMII)
  Kazushige Saga (NII)
  Andreas Savva (Fujitsu)
  Dave Snelling (Fujitsu)
  Jem Treadwell (HP)
  Jay Unger (IBM)

  Minutes: Andreas Savva

* February 28 minutes approval

  Minutes approved with minor correction: Re-check and correct
  confusion with CGS(-WG) and CSG (service). >(done)

* GGF13 preparation

** Naming session
  - Is there a Naming BoF session scheduled? Andrew to check
  - Enough time to prepare charter, etc? Yes.
  - Do we really want to do a full-blown BoF? Not sure if this will
    end up being a WG within GGF. A charter is not a prerequisite for
    doing a BoF.
    - Agreed that Andrew will draft a charter (low overhead and will
      focus discussion better)

** EM Architecture session
  - Ravi cannot make it to GGF13.
  - Andrew thinks he has no conflicts and can lead the EM sessions
    instead.
  - The Naregi project volunteered to do a presentation. They have an
    EMS based implementation.
  - Andrew will call Ravi and check what Ravi had in mind. 

** EM profile session
   - It should be a discussion of what to do. How far the discussion
     can proceed depends on who actually will be attending.

** Other
   - Date profile session is in the same situation as the EM profile
     session.
   - Note takers for each session: session owners should look for
     volunteers and let Hiro know in advance.

  - Security profile status
    - Still not confirmed that it should be a separate document
    - Tom to arrange a call with Frank S and Dave S
    - (Frank cannot make the next call)
    - Takuya to talk with Frank S and check which basic profile
      sessions Frank can attend.

* OGSA BES BoF  

  (Historical note on how BoFs have fared over time.)

  - Proposed agenda: Big EM picture, 'strawman,' discussion on how to
    proceed.

  - There was disagreement with the proposed agenda:
    - No need for a discussion on the large EM picture.
    - If one strawman proposal will be presented then other people
      should also be allowed to present their proposals
  
  - Has the 'strawman' document produced by Andrew G, Mark M, Steven N
    been accepted by the OGSA group? (In other words is it
    'privileged' in some way?)
    - No. It has not been approved. But it does fit in the OGSA EM
      architecture as described in OGSA version 1.0

  - Fitting in the EM architecture is not the main issue for this
    proposed WG. There are people outside OGSA who do not care about
    the general EM factoring but might still want to participate.

  - The EM architecture is given to scope the problem and make sure
    that not too much is taken on. So it might be a good idea to do a
    short presentation to set the overall picture.

  - Proposal that other interested parties should also be allowed to
    present:
    - Two offers on the call: 
      - Globus (Karl C)
      - Naregi (Kazushige S)
    - Probably 10 minutes each is ok; but need to keep a strict
      time-limit.

  - In principle the above proposal was accepted. Also agreed to send
    invitations to particular people who might be interested in giving
    presentations at this BoF.
    - But make sure that they have read the strawman and have some
      idea of what the intended context is.

** Strawman document
   
   - Proposal is that all scheduling and related activities are out of
     scope: the WG starting point is assuming that everything needed
     is available for execution.
   - There is an outside dependency on the 'abstract name'
   - Proposing 4 operations on the "container"
   - Also states and transition diagram for "job"
   - Why not return a WS-Resource?
     - The approach is to try for abstract interfaces as far as
       possible and for as long as possible; leave rendering for
       later.
   - It is a good draft to make sure that everyone understands the
     scope of this activity. (But more detailed discussion on content
     is needed.)
    
   - Asynchronous notifications? A possible topic to discuss.

* Basic profile discussion

  - A short presentation during the first day session, followed by
    tracker review. BP sessions are working sessions

  - WS-policy and WS-SecureConversation references are in the documet
    but are not mentioned. >Delete
  - Also agreed to delete OGSI reference.
  - "Extensibility point specified..." what does it mean if such a
    point is not listed?
    - Reconfirmed that the intention is for the profile to capture ALL
      extensibility points in the referenced specs.
  - Property definition - version (#1246)
    - A version of resource or interface? Is there a structure or not?
      Can we reason about it or not?
    - Many reservations to taking this on; Agreed to push to a higher
      level.
  - (#1248) punt up as a resource management issue
  - 1251: 'known'? Really want to know the ones that are
          there. Discuss further on a call
  - 1252: Read or write property on the resource property. 
          Add text on the resource properties that cannot be
          'set/insert/delete'
  - 1258: Needs further discussion
  - (1254) Also need discussion on whether to mandate query
  - 1294: Tom will put together a timeline as known today
  
  - Open a tracker for the naming profile: Should the basic profile be
    dependent on the naming profile?

  - Which version of the WS-I BP Security was used by Dave S? The
    January working draft.