GlueX Software Meeting
March 8, 2005


Attending: David L. Matthew S. Richard J. Elliott W. Simon T. Paul E.
           Zisis P. Mauricio B Nikolai K. Joachim K. Matt B.
Agenda:

1. Status reports (All)
http://www.jlab.org/Hall-D/software/Software_tasks.php

2. Multiple HDDM versions in repository (All)

3. Meeting Regularity (All)

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

1. We discussed the items on the software task list kept at:
http://www.jlab.org/Hall-D/software/Software_tasks.php

Richard Jones gave a nice update on comparing Geant3 and Geant4.
His notes can be found at: http://zeus.phys.uconn.edu/halld/hdgeant4/

Some interesting points from Richard's report:
1. Geant4 seems to be roughly a factor of 2 to 3 slower than Geant3
2. Geant4 has improvements in handling lower energy (<1MeV) over Geant3.
   This may not affect GlueX much since our calorimeters are less
	sensitive to the low energy particles.
3. Most experiments which have made comparisons have found Geant3 and
   Geant4 to give very consistent results.
4. A Geant4 project to describe geometry in XML format called GDML
   is being developed. This appears very analogous to HDDS, but still
	lacks some features of HDDS (which was based on AGDD). We will watch
	this project as it develops.

There was some discussion as to when we should move to Geant4. The
consensus was that eventually we will want to, but it there is not
a lot of pressure to do so at this point for the same reasons as have 
previously been stated. Namely, that we do not have the reconstruction
software in place yet that could be sensitive to the differences
between Geant3 and Geant4.

We also talked about how to accomodate alignment values in the 
geometry. Richard suggested following the CLAS model of having
a base geometry in files (HDDS for us) and placing the small alignment
corrections in the calibration database. There was general agreement
that this was a good plan.

Richard also mentioned a tracking package he read about in his research
on Geant4 called GeantE which is supposed to be general and could
therefore be useful to us.

Short reports were given by the rest of the attendees on the status
of the other software tasks. Of note:

- David L. said he would provide a set of instructions for downloading,
  compiling, and running the analysis software with HDGeant.

- Paul E. reported Alexander has been working on getting the UPV into
  HDGeant. Alexander will give a report next week.

- GRID tools has been added to the list of Software Tasks.

- Simon Taylor's name was assigned to FDC reconstruction.

- Richard Jones' name was assigned to Tagger reconstruction.


Agenda items 2. and 3. were postoponed until next week.