Present: Alex, Heather, Joanne, Karl, Leon, Marco, Michael, Riccardo, Sasha, Toby, Tracy
Core Report: (Toby) See his report. Most items generated extra discussion.
Gaudi v9: (Heather) Conversion is all done and has the seal of approval. Use the GaudiSys checkout package to pick it up and take a look at the GLAST Gaudi page for all kinds of help and explanations. Gaudi is not yet easily available on the UW terminal server, where currently part of the CMTPATH is not under user control (and points to the old Gaudi). As these minutes were being written Alex installed it on the SLAC Linux system along with required external library updates. It can be found at /afs/slac.stanford.edu/g/glast/ground/releases/GaudiSys_v9r0p4. Note Gaudi v9 is now required with all the G4 based code.
ROOT: (Heather) She has done extensive testing with the new Gaudi, balloon data, etc., and everything looks fine, so this seems like the right time to move forward to ROOT 3.02.07.
Program Constants: (Toby) The xmlGeoDbs package takes care of geometry constants and other constants reasonably closely linked to the instrument. What about constants needed just by a particular algorithm? These also could be kept in xmlGeoDbs files, but this doesn't have the right sort of modularity and flexibility. Another possibility is jobOptions, but this also breaks modularity since each executable using the package would have to know about the associated jobOptions. There was general agreement that an XML file local to the package with an interface like IFile is the best solution.
Release System: (Karl) Everything is up and running: nightly builds (for newly-modified packages satisfying certain other difficult to summarize conditions), documentation updates, and testpack server. Nightly builds output is just ascii for now, but will soon be upgraded to web pages. Still need definition of a "mother package" which will pull in everything of interest. Also Karl needs to be blessed or knighted, or something, in order to use CGI scripts so he can get the Web interface to testpack going.
G4 transition: (Toby, Riccardo) There is plenty already in Toby's report. Additionally,
With regard to tracker digitization, probably want to handle noise, apply thresholds, and calculate TOT outside either of the digi algorithms. [But some forms of noise/fluctuation are already handled by the Bari algorithm. Riccardo will ask Monica et. al. for a document explaining exactly what is done along these lines.] The simple digi algorithm is ready to produce digis for input to tracker reconstructon.
CAL integrating hits have now been made using the xml description with CsI crystals segmented into 12 G4 volumes. They need to be looked at by CAL people, and Sasha is just about ready to do some looking.
(Leon) We need something analagous to pdrApp for the new G4 world which at least makes tracker recon and CAL recon available in the same process.
J. Bogart Last Modified: 04-Aug-2004 15:39:29 -0700