Present: Joanne Bogart, Toby Burnett, Johann Cohen-Tanugi, Marco Frailis, Berrie Giebels, Heather Kelly, Michael Kuss, Sean Robinson, Leon Rochester, Alex Schlessinger, Tracy Usher, Karl Young
Gleam release: (Toby) See his Gleam notes (new features, to-do) for details. Items discussed include:
Failure of test program on Linux. Not yet a critical problem since we are not running production batch jobs, but could become so at any time. In the latest version (v2r2p3) the test program fails after about 150 events rather than just a couple, but the failures look similar. It may be that a bug fix Bill Atwood made, not yet committed, will help.
Need of a proper scheme to limit MC particles. Should be dealt with when Toby visits Udine, if not before.
(Marco) Relational tables for particle/hit relations. The code is done and ready to be committed. You can see an example in the Doxygen output for the test program. The code will go (update: already has gone) into a subfolder of the Event package.
Windows (lack of) optimization. Even when Release is selected, code is built without optimization. It will take some effort to track down exactly how and why CMT is doing this. Available person power to look into this kind of problem is unfortunately limited to some fraction of Toby.
Reduced geometry idiosyncracies. Incomplete identifiers generated when using a top volume other than the full LAT are a stumbling block.
Calibration infrastructure: (Joanne) Design and implementation of metadata table has undergone minor improvements. Interface classes to it are written but need to be tested; similarly for bad strip data (read services only so far), thanks to student help, alas ended. Next step: integrate with Gaudi.
News from France: (Berrie) It is likely he will soon have some help for work on CalRecon. He also might have found someone interested in working on Science Tools down the road.
CAL Code Review: (Heather) Generally speaking the CAL coding effort is looking good. Please take a look at the about-to-be-released draft report. Comments are of course welcome. The next review is tentatively scheduled for July 16, but this is likely to change. Volunteers for reviewers are hereby solicited.
Worth looking into: New userAlg, Replay, allows you to interactively select an algorithm to (re)run from the gui.
previous | minutes index | next |
J. Bogart Last Modified: 01-Jun-2010 15:47:24 -0700