Minutes of the Testbeam Meeting, held Tuesday, December 17, 2002
 =================================================================

 (Present:  ML, NK, TI, MF, TS)


 Up front, so we won't forget when the new year comes around:
 The meeting time will shift slightly from 10:30 to 11:00 every
 Tuesday.



 Software:
 ---------
 Naoko reported that the Atlas repository was updated and collected
 with release 5.1.0. That is the version just before Sven committed
 his timing code.

 The Pedestal Production is done and in castor and will soon be available
 (after checks that it is ok).

 The Run information database is being updated. It is now possible to use
 LArBookkepingSvc, which means that we won't have to specify castor
 directories any more.

 Michel, Rob ans Sven now have commit permission for the Atlas repository.

 The next release (tag collector 5.2.0) is due Jan 13
 the next after   (tag collector 6.0.0) is due Jan 27



 ===> *! ===> It was agreed (by Sven, michel, Naoko and Rob) that
              On Jan 13 (at noon) the hectbmon repository will
              be frozen. Commit permissions will be removed at
              that point and new code will be comitted to the
              Atlas repository from that point on.



 Timing:
 -------
 The second topic of the meeting was Sven Menke's work in Munich.
 Sven implemented code that him and Michel had discussed during the
 LArg week at Cern.
 Sven created a new top algorithm which, on an event-by-event basis
 takes all the cells of all detectors and evaluates their peking time
 (using SimplePolynomial). Then, based on these results, an overall
 time for the event is established.
 The Signal Builder can now use this time.

 Michel has updated his software-schematics diagram accordingly.
 See: http://particle.phys.uvic.ca/~web-atlas/atlas/hec-emec/

 A big leap forward is the fact that the wac's are now all in a file
 and don't need to be specified by the user any more (although it
 is still possible to do). Michel took care of backward-compatibility
 in this file. There are 4 wacs types (depening on trigger type) for runs
 within a given range in the old data, while for the new combined run
 there is one wac for each run.
 Old jobOptions files will still work, new jobOptions files will
 sort out the wac's automatically.

 The next 2 steps in this development will be to make sure that
 the new timing routine indeed works and then to synchronize the timing
with
 the digital filtering weights (which will give us yet another timing
file...).
 Eventually all the timing details will be hidden in timing files
 which will be accessed automatically and the user won't have to worry
 about it any more.

 Sven also proposes an improvement to LArSimplePolynomial which
 supposedly is better at finding small signals (muons).



=========================================================================


 Things to do:

 *  Michel will tell Sven that he (Sven) now has commit permission for
    the Atlas repository.
 *  Naoko was asked by Sven to assign a CLID for LArTBTimingObject
 *  Margret promised to look through Sven's recent emails and summarize
    Sven's progress for our next meeting.

 *  Everbody: Have a good Christmas Holiday! Your task is to rest
    and relax so you will be ready for the new year when it comes!


 Regular meetings will resume next year, Tuesday, Jan. 7 at 11:00 in the
Reading room.