OGSA Teleconference - 26 April 2004
===================================

* Participants

  Dave Berry (NeSC)
  Ian Foster (ANL)
  Andrew Grimshaw (UVa)
  Hiro Kishimoto (Fujitsu)
  Fred Maciel (Hitachi)
  Andreas Savva (Fujitsu)
  Frank Siebenlist (ANL)
  Chris Smith (Platform)
  Ravi Subramaniam (Intel)
  Jem Treadwell (HP)
  Jay Unger (IBM)

  Apologies: Bill Horn

  Minutes: Andreas Savva

* Early Discussion

** Teleconference minutes Apr. 21 - no comments, approved

* May F2F Update
  
  - Will there be a bridge into each subteam room?
    - At the moment there is probably just 1 line (for main room)
    - Could reserve for group that is determined to be highest priority
      (e.g., have people actually call in)
    - In the past people have not called in, even though a bridge was
      set up. 
    - Chris will not be able to attend and wants to join the PE team
      session. He volunteered to set up a call number for the PE
      session if no other arrangements can be made.

  Action: Ian and Chris to coordinate on setting up the bridge for the
          PE session. (Possibly in addition to any other setup.)
  
  - Another suggestion was to publish minutes of each design team
    session before the general session and have a bridge into this
    session only.

  ** If you are attending the F2F please arrange to stay Wednesday
  ** night if possible.

  - Also there might be some extra meetings Monday afternoon (e.g.,
    CMM). Consider joining if you are arriving early on Monday.
    
* GGF11 Update

  - Session scheduling deadline is May 7
  - Hiro has sent proposal to Stacey.

  - Frank would like 1hr to discuss security use cases in a design
    team session
    - Decided to dedicate an entire session (90 minutes)
 
  Action: Hiro to ask Stacey for another session.

* Glossary

  - Jem updated the glossary based on last call's discussion. 

  - Hosting Environment: Ok (with a comment that maybe 'process'
    should also be mentioned.)

  - Definition of Grid Sevice
    - No consensus on whether to define this term in a non-OGSI
      environment. (Ian against, Ravi for).
    - No clear definition either since the properties of the GS cannot
      be listed at the moment.
    - The main concern is that we may be creating confusion by keeping
      this term. 
      - (E.g., WSDM is based on WSRF. Are we eventually going to end
        up saying everything is a GS?)
    - If we do not define GS then what does that say about OGSA.
      - (O GS A or OG SA?)
    - Andrew's perspective is to use/define this term even if it is
      only as a marketing tool.
    - Jay's proposal is to quasi-deprecate the term. The term has too
      much power at the moment so it is probably impossible to
      completely stop using. 
      - (And if we do define the term, perhaps encourage people to
        think of a GS as implementing WSRF in a particular way.)
      - Concern with additional baggage, e.g., OGSI 'otherness'
        implied by the term.
        
    - This sounds more and more like a marketing/political issue
      rather than a technical one.
    
    - Process check since there is no obvious consensus. 
      - Might have to revisit at F2F (...)

   Action: Send opinions to Jem who will summarize.

   - Jem cannot attend next call (Apr 28). 

   Action: Andrew to lead glossary discussion using a couple of PE
           terms from the new glossary.

* Data design team discussion

  - Data team update from Dave Berry.
    - Collecting high level use cases/scenarios
    - Scheduling a long (3 hour) call on Friday (Apr 30)
    - Using a collaboration tool (IBM) to aid online discussions.
    - Maybe have one more call after Apr. 30 and the F2F.
    - Not sure if the design team will be able to have enough text for
      the F2F or the GG11 subsmission.
  - (Andrew) The data team is making good progress and is starting to
    coalesce. It is unlikely to reach the level of PE by GGF11 but
    that's just because there hasn't been enough time.

  - Dave Berry also did a writeup on how this work relates to other
    data groups (e.g., DAIS) and sent it around since there have been
    concerns raised by those groups.
    - Data design team interaction with other groups that might end up
      doing more detailed work may be a typical way of doing things.
    - Would like to get these groups to participate more directly.

  - Some more general concerns with OGSA 'disintegrating' into just
    list of services. 
    - Would also like to look at the problem from a higher level
      (unifying) perspective. E.g., 'Placement' is a term used by both
      Data and PE and would like to look into commonality.
    - Before doing so we need to work at getting some detailed picture
      of the area (current state of data work)
    - In the past assuming OGSI allowed us to make some basic
      architectural assumptions

* AOB

** Provisioning
   
   Action: Jay to send marked up version of provisioning text to Hiro/Andreas/Andrew.

** Other
   
   Action: Jem to put issues with SOAP 1.1 etc raised by Steve Loughran
           in tracker.