Software Coordinating Committee Conference Call
		       December 11, 2003 3:00 PM EST
			  Recorder: C. DeTar

Present:

  Eric Nielsen, Holmgren, Simone, DeTar, Osborn, Watson, Mendes, Brower  
      Edwards, Mawhinney, Pochinski

Absent:
  
  Neff, Petrov, 

** action items
=====================================================================

    #0.QCDOC

        Mawhinney: 128 nodes working now.  Naive staggered hmc
          running now - Chulwoo running comparative tests.
        Brower: Any progress on standards for interfacing
          with level 3 routines.
        Mawhinney: Not finished since Eric's visit.
        Brower:  Should we ask Eric to propose an interface?
        Mawhinney: They should finish what they started.
        DeTar: Are things set for Eric to work remotely?
        Mawhinney: Yes.  Need more time to develop the programming
           model.  Perhaps early January to get him involved again.
        Brower: We should alert Eric to be ready.

     #1. Progress on report from Meta Data Working Group.

	(Robert, James, Carleton: Can one or more of you summarize
	Last Fridays ILDG workshop?  Please send pointers to any
	slides you wish to share with us.

        DeTar: Gave a summary.  See Chris Maynard's slides and
          DeTar/Edwards presentation
           http://www.ph.ed.ac.uk/~cmaynard/QCDML.ppt
           http://www.physics.utah.edu/~detar/SciDAC_comment_20031205rev.ppt

     #2. Chip, would like to present your schedule for the Middle
            Ware working group? 

        Watson: Gave summary of architecture.  See Watson presentation:
           http://www.ph.ed.ac.uk/~bj/ILDG-middleware-chip.ppt
        DeTar: How does the coordination of national implementations
           work.
        Watson:  The queries will be broadcast to all the local
           metadata catalogs and the results merged.  There
           have to be standards for interfaces.
        Nielson: Still a lot of details to be decided.
           e.g. central query service could simply hand clients
           the list of local services and let the client decide
           which ones to query.
        Watson: Went through the timeline
        Brower: Is the Middleware Group assigning tasks now?
        Watson: Yes.  We will know soon what our manpower requirements ar.

        Brower: What kinds of access control should an international 
           resource have?
        Watson: Access controls to be decided next Fall.

        Mawhinney: Does this really relate to SciDAC?
        Brower: It touches on what we are doing.

        Further discussion led to the following agreement about
        how the middleware relates to SciDAC software;

        Applications don't interact with the archive -- only with
        local files.  The ensemble XML will be created locally and
        uploaded as a file.

        DeTar: I assume this could be either from our application code
          or from tools provided by the Middleware Group for generating
          schema compliant documents.
        Watson: Yes
        DeTar: And how would we edit the ensemble metadata after it is
          uploaded.  For example, as a given ensemble is extended, the
          management information needs updating.  And as we obtain
          references to derived data, we should be adding them.
        Watson: Yes, we need to plan for such use cases.

     #3. SciDAC Renewal Proposal

**      Brower: I will post the current (nearly final) draft of the
          renewal proposal.

     #4. Set priorities for 2004. 

=====================================================================
Committee conference concluded at 4:35 PM EST.  
Next plenary call December 18 at 3 PM EST.