Minutes of the Testbeam Meeting, held Wednesday May 21, 2003
 =================================================================

 (Present:  NK, TI, MF, TS, RK, IG)


 TBRootAna:
 ----------

     Ian reported that he has completely reworked the package so that
     it now works with the standard gnu tools, which take care of
     automatic dependencies during the build process. This should
     make things a lot more maintainable.

     He has also started to implement the new Ntuple data structure
     and a preliminary version for the 2002 data should be ready
     by the end of the week.


 Geometry files:
 ---------------

     Richard and Margret have joined forces to create EMEC and HEC
     geometry files that will include the actual z-position of the
     individual cells. The files are essentially ready now.
     (Thanks should also go to Fares Djama who provided us with
     the eta-dependent z-coordinates of the S1/S2 and S2/S3 boundaries
     of the EMEC.)

 Geant4:
 -------

     Naoko reported that Pavol's Geant4 package compiles and runs
     on the fates.
     It was suggested that it should be put into a common place,
     somewhere in the Atlas software directory structure.


 Tayfun
 ------
     has looked at some 119 GeV electron data. He is trying to
     find the cells with the largest energy per event.
     He is currently working on how to deal with noisy channels,
     which are a problem in events that don't see significant
     total energy deposits. At the moment it is unclear
     whether these events are random, muon or something else.


 Tamara
 ------
     is working herself through the ROOT tutorial to get reacquainted
     with the important analysis tools.


 Problems - Questions - Answers
 ------------------------------

 Tayfun noticed that h2root isn't working any more on the fate,
        however at Cern it works.
        Ian mentioned that what is needed is likely to specify
        the new root (/net/cern/root-3.05). This however only
        works on fate3 at the moment.

 Richard was wondering whether there is anything in the way
        of some large Ntuple production.
        Naoko said that at the moment it is just the out-of-phase
        events; a fix for that is committed but not tagged yet.