Software Coordinating Committee Conference Call
			   XML Subcommittee
		       June 2, 2003 3:00 PM EDT
			  Recorder: C. DeTar

Present:
  Osborn, DeTar, Simone, Mawhinney

=====================================================================

  Edwards
    
    I sent around a proposal to this subcommittee, expanding on
    our metadata list from Thursday's plenary meeting,
    merging with Yoshie's proposal, and attempting a similar
    streamlining of the algorithms description.
    (See attached)

  Subcommittee consensus:

    The proposal goes in the right direction.  The following improvements
    were proposed:

    Algorithms description

      The algorithms section should contain keywords for describing the
      updating algorithm together with references to the original publications
      and a reference to a companion document giving specific details
      that may have been omitted from the publication.  A reference to the
      code, including version number and compilation details would also
      be good.  The goals here are (1) making it possible for a
      researcher to extend the ensemble of configurations, either with published 
      code or with his/her own code.  (2) providing an audit trail in case
      problems with a particular code version are discovered later.

    Glossary metadata

      We discussed having a glossary of XML files that expand on unique
      keywords for algorithms.  These could include the detail that
      UKQCD needs.  So the dataset XML might mention "Asqtad" and the
      glossary could list all the terms and couplings.

    Some more technical details;

    File XML

      We probably don't need the header version number.

    Convenience function for 

      There should be a convenience function to initiate the creation
      of the file XML string with the archive-required fields.

      There should also be a convenience function for writing a complete
      gauge field.

    BinX and private record XML

      With QDP-IO we will doubtless need to parse the BinX to be able to
      insert the correct data for arrays of QDP fields.  We will
      investigate whether it is possible to insert our own labels for
      field types.  We may then be able to dispense with the private
      record XML.

** Robert will revise his proposal and circulate it to the full
   committee.  We are hoping for an early reply to Yoshie's committee.


----------------------------------------------------------------------
Date: Mon, 02 Jun 2003 12:59:29 -0400
From: Robert Edwards <edwards@jlab.org>
Organization: Jefferson Lab
X-Accept-Language: en-us, en
To: Robert Mawhinney <rdm@phys.columbia.edu>,
        Carleton Detar
 <detar@physics.utah.edu>, Jim Simone <simone@fnal.gov>,
        James Osborn
 <osborn@physics.utah.edu>
Subject: proposal for xml of gauge field

This is a multi-part message in MIME format.
--------------010601010400010102080903
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Please find attached a proposed xml instance for our gauge fields.
I've blended our own decisions plus some from the meta-data working
group.

            Robert

-- 
Robert G. Edwards                          phone: (757) 269 7737
Theory Group, MS 12H2                      fax:   (757) 269 7002
Jefferson Lab
12000 Jefferson Avenue                     edwards@jlab.org
Newport News, Virgina  23606, USA          http://www.jlab.org/~edwards


--------------010601010400010102080903
Content-Type: text/xml;
 name="gauge_schema_scidac.xml"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline;
 filename="gauge_schema_scidac.xml"

<!-- File XML -->
<file>
  <hdr_version>1</Hdr_version>
  <creationDate>June 2, 2002</creationDate>
  <ensembleID>012345</ensemblID>
  <sequenceNumber>55</sequenceNumber> <!-- monotonic -->
  <dimension>
    <int>16</int>
    <int>16</int>
    <int>16</int>
    <int>32</int>
  </dimension>
</file>

<!-- Record XML -->
<record>
  <datatype>4D_SU3_GAUGE</datatype>  <!-- a simple key -->
  <linkTrace>-0.53</linkTrace>  <!-- gauge configs only -->
  <plaquette>0.67</plaquette>     <!-- gauge configs only -->
</record>

<!-- BinX record -->
<dataset>
  <definitions>
    <typeDef typeName="complexFloat">   <!-- Either Float or Double allowed -->
      <struct>
        <ieeeFloat-32 varName="Real"/>
        <ieeeFloat-32 varName="Imag"/>
      </struct>
    </typeDef>
  .....
    <file src="sample.config#3" byteOrder="bigEndian">
</dataset>
 
<!-- Record checksum -->
<!-- recordered as a simple binary (long ?) int -->
checksum  

<!-- Dataset XML: a distinct file -->
<gaugeConfiguration>
  <ensembleLabel>Some user supplied string</ensembleLabel>
  <ensembleID>012345</ensembleID>          <!-- same symbol as in file xml -->
  <archiveDate>July 30, 2300</archiveData> <!-- date in future when configs archived -->
  <references>A PRD string ref.</references>  <!-- Publication -->
  <markovStep>
    <physics> ..... </physics>      <!-- See below -->
    <algorithm> ..... </algorithm>  <!-- See below --> 
  </markovStep>
</gaugeConfiguration>


<!-- Physics part: modified from Yoshie's proposal -->
<physics>
  <size>
    <int>16</int>  
    <int>16</int>
    <int>16</int>
    <int>32</int>
  </size>
  <gluon>
    <plaquette_gauge_action>
      <plaquette_operator>
        <coupling>
          <coupling_name>beta</coupling_name>
          <coupling_value>5.2</coupling_value>
        </coupling>
      </plaquette_operator>
    </plaquette_gauge_action>
    <boundary_condition>
       <string>periodic</string>
       <string>periodic</string>
       <string>periodic</string>
       <string>antiperiodic</string>
    </boundary_condition>
  </gluon>
  <guark>
    <sw_fermi_action>
      <wilson_fermi_operator>
        <coupling>             <!-- Example of 2+1 simulation -->
          <coupling_name>kappa_1</coupling_name>
          <coupling_value>0.1350</coupling_value>
        </coupling>
        <coupling>
          <coupling_name>kappa_2</coupling_name>
          <coupling_value>0.1375</coupling_value>
        </coupling>
      </wilson_fermi_operator>
      <sw_fermi_operator>
        <coupling>
          <coupling_name>c_sw</coupling_name>
          <coupling_value>2.02</coupling_value>
        </coupling>
      </sw_fermi_operator>
    </sw_fermi_action>
    <boundary_condition>
       <string>periodic</string>
       <string>periodic</string>
       <string>periodic</string>
       <string>antiperiodic</string>
    </boundary_condition>
  </guark>
</physics>


<!-- Algorithm -- >
<algorithm>
  <hmc_molecular_dynamics>
    <number_of_updates>160</number_of_updates>
    <time_step>0.00625</time_step>
    <residual>1.0e-7</residual>
  </hmc_molecular_dynamics>
</algorithm>

--------------010601010400010102080903--

----------------------------------------------------------------------

Subcommittee conference concluded at 5:15 PM EDT.  
Next plenary call June 12 at 3 PM EDT.