OGSA Teleconference - 3 August 2006
===================================

* Participants

  Jem Treadwell (HP)
  Jun Tatemura (NEC)
  Ellen Stokes (IBM)
  Dave Snelling (Fujitsu)
  Andreas Savva (Fujitsu)
  Joel Replogle  (OGF)
  Darren Pulsipher (EMC)
  Dejan Milojicic (HP)
  Tom Maguire (EMC)
  Hiro Kishimoto (Fujitsu)
  Michel Drescher (Fujitsu)
  
  Minutes: Andreas Savva

* Summary of actions

  AI-0803a: Hiro to look into an option for a second EMS/Data call
            on Sep 7. Tentative; needs confirmation with likely
            participants.

  AI-0803b: Michel to revise and upload Joint CDDLM minutes (Done)

  AI-0803c: Ellen to book the IBM facility on Thursday afternoon
  AI-0803d: Hiro to divide GGF18 B2B F2F survey into three periods:
            Thursday afternoon, Friday morning and afternoon.
  AI-0803e: Jun will provide an example of the CDL mechanism and why
            using ID/IDREF is not a good idea.
  AI-0803f: Jun will update the CDL document to remove the Environment
            variables

* July 31 minutes approved with no changes

* Review of action items

  [Only items with updates are listed below. '+' marks added text.]

  - Ellen to send an email to Chris Jordan with the information
    needed to add these two documents to the OGSA Roadmap
    + Information sent. Done.

  - Hiro schedule a joint call with OGSA-Data to explain the
    Information model work before GGF18.
    - This could be one agenda item on the joing EMS/Data call
      already being schedule.
    + Joint call between EMS/DATA & Info Model is scheduled for August
      17. Done.

  AI-0803a: Hiro to look into an option for a second EMS/Data call on
            Sep 7. Tentative; needs confirmation with likely
            participants.

  - Jun to feed back proposals 1a and 1b to the CDDLM Working Group,
    i.e. pass back the minutes to the WG. Ask Michel in case of any
    questions related to these proposals.
    + Discussed during this call. Closed

  - BES-WG chairs to ask the data group to define the
    filesystems as their work product
    + Hiro sent actions to WG chairs. Confirm that this is so and
      close.
    + Put on the agenda of the joint EMS/Data call.

* July 2006 F2F minutes approval

** 20060718-pm-2 (GFS) - approved with no changes

** 20060719pm-2 (Joint Mgmt Committee) - approved with no changes.
   - Confirmed that there are no concrete actions on OGSA-WG at this
     time.

** 20060720-am-1 (Joint CDDLM) - approved with changes
   - Change XPath to XQuery
   - Modily statement on "removing environment variables from CDDLM"
     to removing from Posix component. (Proposal 2 correctly states
     this.)
   
   AI-0803b: Michel to revise and upload Joint CDDLM minutes (Done)
   
** 20060720-am-2 (EMS Scenarios) - approved with no changes
** 20060720-am-3 (G.Fox comments) - approved with no changes

** Roadmap minutes are not available (postponed)

* GGF18 update 

** B2B F2F 

  GGF18 is probably finishing at 1pm on Thursday. Agreed to try and
  start the interim meeting from Thursday afternoon since many people
  are likely to want to leave early on Friday.

  Tentative start time: 2:30pm on Thursday

  AI-0803c: Ellen to book the IBM facility on Thursday afternoon
  AI-0803d: Hiro to divide GGF18 B2B F2F survey into three periods:
            Thursday afternoon, Friday morning and afternoon.

** GGF18 sessions

   - BES clashes with EMS session. Asked Joel to try and resolve it by
     moving one to a different slot. There are many dependencies,
     however.

* Joint CDDLM discussion

  Discussed proposals from the F2F minutes (reproduced below and
  marked with '>')

> PROPOSAL 1a: Use XPath 2.0 as an interface between the CDL document
>              and the CDL components

  - Clarified that this should be XQuery not XPath. 
  - Agreed that it would be nice to do but it is probably a feature
    for the next version of CDL. It is one topic to discuss at GG18 as
    part of the future direction of the CDDLM-WG

  (See also AI-0803b)

> PROPOSAL 1b: Use xsd:ID/xsd:IDREF to allow relations between elements in
>              CDL documents

  - Jun thinks that this is not a good idea because ID must be unique
    in the document and CDL inheritance may break this guarantee by
    creating duplicate IDs in the resulting document. CDL has a
    mechanism based on non-unique ids.

  AI-0803e: Jun will provide an example of the CDL mechanism and why
            using ID/IDREF is not a good idea.

> * Proposals 1a and 1b are recommendations for the further
>   development of CDDLM

  Restated that both these proposals are for post-version 1.0 of CDL.

> 
> PROPOSAL 2: Eliminate the EnvironmentVariable concept from the
>             PosixComponent (CDL plugin)
> 
> * Proposal 2 is specific to the OGSA EMS scenarios document.

  Agreed that this is a reasonable resolution for resolving the
  problem with the PosixComponent and is sufficient for the EMS
  Scenarios document.

  AI-0803f: Jun will update the CDL document to remove the Environment
            variables

  (See also AI-0803b)

  - A normative definition of PosixComponent 
    - At the moment there is no such definition; there are only some
      examples.
    - Agreed that a normative definition is necessary and that it is
      not part of CDL version 1.0. Also it is probably too late to
      leave it for CDL version 2.0---since it is not scheduled yet.
     - Tentative agreement that this should be a work item for
       CDDLM-WG between these two document release
     - Dejan and Jun with discuss this in the CDDLM-WG and seek
       approval as a work item after CDL 1.0.