***********************************
   minutes of EW task force - june 6
  ***********************************
 
 
1. On the DSTs produced by H.Videau for 90-91-92 data (A.Venturi):
 
   - the 91 DSTs have been produced BEFORE the last
     POT to POT reprocessing of september 93
 
     For these data a class 24 selection was done recently by
     Gerardo and this is the place where the photons with
     negative energies have been found (tape AT0087: 91 off-peak).
     In fact it was later found out
     (see my mail of this morning) that the sign of many
     other real numbers is inverted w.r.t. the original data.
     This is maybe linked with a change of library on shift9
     which took place a couple of weeks ago.
 
                  BE CAREFUL!
 
     Check the content of the banks if you have used shift9 in the
     last days. Follow the ALEPH news.
 
2.   The 1992 data should be class 24 but since the selection was
     private, the EDIR class bit map does not contain the info.
     If the class 24 is used, no events are selected.
 
     The 1992 data tapes contain two duplicate runs: 17081 and 17546.
     Despite of this, a comparison with the official DST event
     directories, class 15, gives less events than in the official
     DSTs.--> to be understood
 
     This cross-check gives consistent results for 90 and 91 peak
     data.
 
3.   Bhabhas: a cross-check was done between bhabhas selected
     in the standard way and those selected by Henri-JC.
     These were p-2 93 data. 22 events are found by Henri-JC
     and not by Lee/Pere (on a total of 4617 events).
     The comparison has to be completed. Are there taus in those
     22 events? Are there events found by Lee/Pere and not by Henri/JC?
     2 runs are selected by scanbook but do not appear in Henri's list.
     --> to be understood
     Also the comparison should be extended to 95 data.
 
     The way in which errors are attributed to the efficiency
     should be reconsidered (error=0 if all events are selected
     in a certain costheta bin).
 
4.   Taus: all n-tuples for all the years have been produced,
     in principle. The effect of the shift 9 problem
     should be studied. Are the ntuples correct?
 
5.   qqbar: TPC and Calo selections are being cross-checked
     one vs the other.
 
6.   Lumi: it is expected to have preliminary syst. errors
     and an update of the Scanbook luminosities by the end
     of the week.
 
7.   It is possible to produce NOW a MC for 90 data (GALEPH 241
     restored by Florence) with a VDET
     description matching the 90 VDET .
 
8.   It was decided that clearly priority one for Warsaw is
     to complete the 95 cross-sections keeping the systematic
     errors as they were before, since the studies are not
     yet finished. The cross-sections before 95 will most likely
     not be changed for Warsaw.
     The work being done now for the previous years will be put
     on a paper to be prepared this summer.
 
     For the asymmetries the situation is more complicated.
     We cannot give a preliminary update of all the asymmetries
     since the numbers up to 92 have been published.
     To only update 95 is also unsatisfactory since the present
     results 90-93 are certainly much better
     than those that we have so far shown to the outside world.
     There was no clear conclusion on this point.
 
     A status report on all these activities will be given next week
     (13 june) at the th.meeting.
     Plots from the various selections and analyses supporting
     what we are doing are more than welcome.