Core Issues 15 August 2001

Previous status report; 

Infrastructure described in software PDR report, presentation.

pdrApp v6 released: linux problems now solved, and 5M background events run for PDR studies. (See log., and package summary)

• All attributed to attempt to use Gaudi's random number service inappropriately. (bad hack)
• Conclusion: avoid the LD_LIBRARY_PATH mechanism; environment variable loading works fine (last status report has a summary how it works.)

Final tuning for PDR

• New display setup (due v6r2):
  • Detector display moved to a new Algorithm, accessible for irf readback. (DetDisplay)
  • Event display reps for charged and neutrals (Leon); put into a new Algorithm, GismoEventDisplay. Cleverly plots charged tracks after neutral
  • Employ the subAlgorithm mechanism. Both DetDisplay and GismoEventDisplay are started as subAlgorithms of GismoGenerator. But only if the GUI service is available.

     

    See code fragment from GismoGenerator:

     

    
    

  // get the display sub algorithm if( createSubAlgorithm("DetDisplay", "DetDisplay", m_detDisplay).isFailure() ) { log << MSG::ERROR << " could not set up DetDisplay " << endreq; return StatusCode::FAILURE; }

•  

• Fix to display depth of event for irf playback.

• Improvements in flux package allowing more flexibility in defining background spectra.

CMT cvs checkout problem

It appears that our checkout package scheme will not properly retrieve the set of packages, when using CMT's cvscheckout command. If a package, say TrkRecon v2r4, is requested, but v2r5 exists in the repository, CMT asks CVS to checkout the latter! (VCMT is not so "smart".) This is on the table to resolve.

Linux setup at SLAC

The current group .cshrc is broken, needs to be available to define the setup for a pre-built release, with appropriate documentation.

T. Burnett