Core Minutes 9/4/2007

 

Present: David Chamont, Jim Chiang, Berrie Giebels, Navid Golpayegani, Heather Kelly, Michael Kuss, Bryson Lee, Chuck Patterson, Igor Pavlin, Leon Rochester, Tracy Usher, Eric Winter

 

·         News from Heather:

GR HEAD 1.964, testGleam terminates oddly with an exception in OBF. From poking around in Visual Studio, she has the impression that CAL code is implicated, at least in some tracebacks. She has run Valgrind, and will be looking at the output shortly.

(Leon sees something similar in GRv12r2… except that it occurs after ~40K 10-GeV muon events. So maybe the bug doesn't specifically relate to the latest OBF changes after all.)

There was a problem in the astro package; Tracy worked his customary magic on it. Should Toby make a new astro tag? For now GR is using astro v2r12 instead.

There was a chat with Riccardo on Thursday (see: https://confluence.slac.stanford.edu/display/SAS/Meeting+August%2C+2007 ).

Riccardo's plan is to tag the HEAD of MRStudio after completing some updates requested by Joanne.  This next version of MRStudio will be a beta version released to developers for testing.  Riccardo will then complete updates of MRStudio using CMT including support for VS 2005, then freeze it and move to SCONS world.

Richard reports to Heather that we'll get all the hardware we requested, partly from an over-purchase by BaBar!  We need to start thinking about interleave and event collections. [Haven't we been all along?]

 

·         Navid:

SCONS -- Overwriting of packages is a work-in-progress.

Currently, after compilation, the files are assembled and a(n) SCONS script is made(?); the next step is to add an option allowing the user to over-write the files.

The interview with the potential hire at Goddard went well. They are expecting to make an offer, pending reference checks.

 

·         David:

Event Collections -- He'll be sending an email summarizing progress. A big step forward is sorting out the semantics. The old "event collection" or "meta root data" should be called something like "CompositeEventList," to correspond to ROOT usage.

With the new system, the skimmer could generate an event list, rather than assemble the actual file of events.

 

[At ~8:20 I lost audio; apparently the meeting ended at ~8:22.]

 

  -- Leon Rochester