Core Minutes 9/19/2006
Present: Joanne Bogart, Toby Burnett, James Chiang,
Tom Glanzman, Navid Golpayegani, Igor Pavlin, James Peachey,
Leon Rochester
- Data server back-end: (Tom)
Not much news since last week. (David Chamont is at
a conference this week.) Tom and David would like to put
it into production in backwards-compatibility mode.
[Update from Tom, post-meeting: turns out maintaining
backwards-compatibility would be bad for the development schedule, so the new
plan is to go to a totally new back-end, combined with the new front-end that
Igor is working on. This won't happen before mid-October.]
- OS porting: (Navid)
No change since last week. There are RHEL4 machines
in the LSF pool, but no way for us to use them as yet.
As for Visual Studio 2005: he has no access to it;
it's not installed anywhere
at SLAC.
- Source ids (Jim) He has tagged code
which takes care of a potential mismatch of source ids (and so
event ids) in obsSim and Gleam—as
long as definitions in job options files match those in the text file
obsSim uses.
- IRFs (Toby, Jim) We need a new set
for current classification trees. Must accommodate
changes to point-spread function.
- Beamtest (Leon) This is not the forum
for an extended report, but it appears that all the effort that went
into the Beamtest was well worth it. There is much to be learned
from the data. For example
- There is substantially more backsplash in the real data than in
the MC.
- Cluster widths are bigger in the data than in MC
- RM Windows problem (Navid) The source of the
difficulty was a machine which was put into the LSF pool without being properly
configured. It didn't have all our stuff loaded onto it. Navid does
not have direct control over what is put into this pool, nor is he able to
select a subset of machines without intervention from SCS.
J. Bogart, Last Modified:
01-Jun-2010 15:47:41 -0700
previous
