Simulation meeting minutes, 8 November 2004

• Discussion items:
  • Collaboration meeting:
    • Do we want a plenary session? Not much new and exciting in simulation except for PEP/MDI which will have its own session.
    • How about a parallel session? Possibly, to get feedback from other groups (however these have not been well-attended in the past).
  • Summary of SLAC Geant4 Users meeting (Dennis)
    • a few useful tools to be added to version 7.0 such as retrieval of cross sections for a given particle and energy
    • singleton particle classes will be removed - this will cause some BaBar code to be re-written
    • Geant4 will migrate to CLHEP 1.9 or 2.0. Either one will require a significant BaBar effort in order to be able to use Geant4 7.0 and beyond.
  • Testing MOOSE in 16.0.0 (Dennis)
    • See the HyperNews item from Doug Smith: Testing Moose in 16.0.0
    • Messages from GentleGenocide" have been around for a long time (since SP3). Most of them can be turned off.
    • Messages from DRC (bizarre tracks) are due to introduction of recoiling nuclei in hadronic models - Dave Aston will remove the warning.
    • other problems like: DchOprTrickleLER::DchOprTrickleMon.cc(315) DsToPhiPiPiPi_Loose::CompSelectCand.cc and DcxTrackFinder::TrkHelixFitter.cc are not sim tags
    • the seg fault with GarbageCollectorThread is a ROOT problem
    • 5% of jobs crashing may be a sim problem, but Bogus has not yet been tested in 16.0.0. Plan to start that soon.
  
  
• Core issues:
  • Cross-platform reproducibility (Dennis)
    • There are few-percent differences in almost all QaSimHist histograms when comparing RH 7.2 and RHEL. This includes the generator histograms. Abi will look into the generator part of this.
    • There are also large differences between RH7.2 and Sun
  
  
• Generators (Abi)
  • See the HyperNews posting: Summary of changes from generators meeting
  • need to suppress phi production in jetset - too many kaons being produced
  
  
• Subsystems:
  • PEP/MDI/Backgrounds (Bill)
    • Have successfully written n-tuples from a long job of HER data. However there is a ZEBRA crash when trying to read the n-tuples. They are too large (1 Gbyte) to be read by the current version of PAW.
    • A set of LER rays is now ready to use
    • Mario made a new tag of the beamline geometry. This allows Geant4 to close the geometry with a smartless parameter of 0.2 (it would not close earlier with pepSmartless set to 0.6). However, there is still a crash after two events have been run.
  • SVT:
  • DCH (Jeff)
    • The final SP7 threshold function and new crosstalk model have been released and sent out to various groups for testing. If all goes well, new conditions will be loaded.
  • DRC:
  • EMC:
  • IFR (Sasha)
    • LSTs: software upgrade: now for a single streamer in the LST, there will be only one GHit in a phi-view. In this past multiple hits were allowed. There was also a bug fixed which caused double counting in the LSTs. No multiplicity table are used now.
    • Nothing remains to be done with the LST software until cosmic ray data can be taken in order to get background frames for mixing.
    • No alignment currently exists in the code. For that, dimuon data (or maybe cosmics) are needed.
  • Trigger:
  
  
• AOB
  • Next meeting: 22 November 2004