Minutes of the UVic testbeam meeting on May 29 :
 ================================================

 - Margret has two ntuples of electron data ready for
   conversion by Ian and for testing out. (40 GeV electrons, one
   ntuple in ADC, the other in nA)
   She had done the pedestal computation over 5 time slices 
   with nsigma=2. However it should have been done over the first
   time slice only and nsigma=3. So the ntuples will have to be 
   redone. Still - the first two should be ok for testing out.

 - Ian showed a quick plot produced in ROOT of some 120 GeV pion data, 
   Tamara had done the same thing, proving that the tbrootana package
   is up and running for analyzing Ntuples.

 - There now is a need for two special accounts:
   - One that will hold Ians CVS repository,
   - another for storage of testbeam Ntuples at UVic.
   Since Ian's package will likely serve more purposes than
   just the HEC testbeam, it was felt that is should be kept
   seperate from testbeam-specific accounts.
   Margret talked to jan after the meeting and Jan agreed with that.

   As a possible acount name came up:    cvs-atlas   
   No suggestions for the ntuple storage yet... (maybe data-atlas?)

   => Please give me your comments and/or ideas on this!
      If I don't get any comments, I will arrange for two
      accounts (cvs-atlas, data-atlas) by the end of this week.


 - Tayfun reported that he is able to do fits on W pair angular
   distributions in Root.
   He finds that there is a 2% difference in the results
   compared to using fortran(minuit).  
   One of the problems he ran into is how (in C++) to implement
   a number in a formula to double precision without declaring it first.
   Ian recommended to cast it, but nobody was really sure what would
   work and a few tests were suggested....