Core Minutes 6/6/2006
Present: Joanne Bogart,
Toby Burnett,
Richard Dubois, Dan Flath,
Riccardo Giannitrapani, Tom Glanzman, Navid Golpayegani,
Heather Kelly, David Landriu, Michael Kuss, Bryson Lee,
Igor Pavlin, Leon Rochester, Tracy Usher
- MRStudio: (Riccardo)
The necessary command-line functions — checkout, configure, make, clean
— are now working on both Linux and Windows. Unless he hears otherwise
from users,
Riccardo believes this phase of the work is done. There is also online help.
Just type
MRStudio help
or
MRStudio help command
- Processing: (Richard)
We're at 2.5B background events with GR v9r6. Did 1-day extragalactic diffuse.
Now looking at when to do 18M allGamma (to make 20) and then 2.5B more
background.
- Beamtest:
(Michael)
- Francesco was making a tag v5r0p3 of beamtest06. This tag is
capable of simulating SPS. He started to do simulations. Unfortunately he
is at the
Swift
conference at Venice this week, hence I don't know if he
succeeded or not.
- Heather fixed LdfConverter and updated a
bunch of other packages
too. Last night, she cleaned up the readigi_runrecon, and Joanne fixed a
problem with xmlGeoDbs.
Today, I could convert a full LAT data file to digi and
furthermore to recon.
- For CU runs, this works only if I pretend that the data are coming from
LAT. Seemingly, for one device there is already an entry in the data
base, but parameters are missing. Hence, conversion terminates during
loading. We can work around this by:
-
pretending it's LAT (using wrong calibs), and then generating the
initial set of calibs.
- filling the database with dummy numbers before doing conversion and recon.
(Richard) Nicola Omodei is handling Recon. Work is also ongoing to support
the ancillary data. Need to configure a Pipeline task to automatically
handle CU data.
- GlastRelease: (Toby) Heather has
made a GR tag (v9r7) which differs from v9r6 only in LDF stuff. We need
an end-to-end test for it. (Heather) It can't be a simple unit test for
one of the LDF packages as they stand since none of them has such an
all-encompassing set of use statements. For EM, this function is
handled by the LatIntegration package. (Toby)
Another way to do this is with a contained test package, as in
FluxSvc/test_FluxSvc.
- Misc:
- Toby reminded us of the decision to support unit test subpackages,
contained in the package to be tested. Navid will look into ramifications for
RM.
- Pat's Confluence area devoted to algorithms allowed anyone
to write to it and anyone—well, someone— did scribble
all over it. We won't do that again.
Thank you, speakers,
for your help in compiling (and I do mean compiling) these minutes.
EVO was not very well behaved today: audio was choppy, especially when
speaker and listener were on different continents. [ed.]
J. Bogart, Last Modified:
01-Jun-2010 15:45:12 -0700
previous
