Software Coordinating Committee Conference Call
		       February 12, 2004 3:00 PM EST
			  Recorder: C. DeTar

Present:

   DeTar, Osborn, Simone, Holmgren, Gregory, Brower, Edwards, 
   Neff, Mendes, Mawhinney, 

Absent:

   Pochinski, Watson, Petrov, Jun, 


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


1. HEPAP meeting

   Brower:
    Only nonexperimental presentation.  Thorough discussion.  Staffin
    was there the whole time.  Good questions for Bob's presentation.
    Committee was well prepared.  Potential improvement in unitarity
    triangle figured centrally.  Second day discussed the writing of
    the report.  Each person on the ocmmittee praised the need for
    lattice calculations.  Indications: '05 few $M

    Concerning FY06 budget: Roy Whitney, Vicky White, Tom Shlagel
    suggested an OMB 300 form for IT investments.  Labs will do it for
    us, using an outside contractor.

  DeTar:
    The Executive Committee and especially Bob deserves our
    praise for an excellent presentation.

  Brower:
    And Steve Sharpe for his analysis of matrix elements.
 
  Mawhinney:
    First day, the tone was serious in face of declining budgets.

  Brower:
    Role of international collaboration: need to have something
    to share in order to share.

2. QCDML

   Simone: I sent our response Tuesday.  So far there have been no
      reactions.

3. LIME format

   DeTar: Are we ready to offer LIME and our file format to ILDG?
   Edwards: So far it has been working for me.  I think we are ready.
   Simone: And sending it out might help ILDG gel.  We should offer
     to converge onto a common format.
** DeTar: I'll put together a description.
   Edwards: And an explanation why we abandoned DIME.

4. QCDOC I/O

   DeTar: About QCDOC I/O.  If we knew what was needed for QCDOC, we
     could consider building in the hooks in the current QIO package.
     That might save some duplication of software effort.
   Mawhinney: We had not placed I/O high on our list, but should
     begin thinking about it.  I will speak with Kostya.
   DeTar: Let me know when you have someone assigned to the project.
  

5. QMP

   Mawhinney:
     Pushing Chulwoo to proceed. Is the issue of handling all-to-all
     communications on grid clusters settled?  Or have we backed
     away from the original spec?

   Edwards:
     No, on clusters we have implemented the complete spec.  We (at
     Balint's urging) have added some crutches - QMP_route for QCDOC,
     but we regarded this as a temporary expedient until the
     implementation is mature.  We never regarded this as correct.  We
     just needed it for debugging QCDOC software.

   Brower:
     Is Balint communicating with Chulwoo?

   Edwards:
     Yes, very much so.

6. MILC on QCDOC

   Gregory: 
     I am working on it and communicating with Chulwoo and Boyle.  Ran
     into a few snags.  Something about DDR overflow.  Probably an
     OS bug.  Peter and Chulwoo are on the problem.
   Mawhinney:
     I don't want you to be stalled.  If you haven't heard back by
     tomorrow, let me know.

=====================================================================
Committee conference concluded at 4:00 PM EST.  
Next plenary call February 19 at 3 PM EST.