Core Minutes 4/25/2006
Present: Joanne Bogart,
Toby Burnett, David Chamont, Eric Charles,
Jim Chiang, Johann Cohen-Tanugi, Seth Digel, Richard Dubois, Warren Focke,
Riccardo Giannitrapani, Tom Glanzman, Navid Golpayegani,
Heather Kelly,
Chuck Patterson, Igor Pavlin, James Peachey, Leon Rochester, Tom Stephens,
Tracy Usher
- MRStudio: (Riccardo) Other than a couple
small problems on Linux which should be solved soon, the batch part of
MRStudio is working; it's available for Windows beta testers. (Richard) Any reason
this is not in CVS? (Riccardo) No reason. He will check it in.
The GUI part is frozen until the above problems are resolved, but it
is only a wrapper in any case. The design of MRStudio, with GUI strictly
separated from internals, means that a scripting interface is available
"for free", but will anyone care to use it? (Richard) Yes, indeed. The
hope is that ultimately MRStudio will be
invoked by RM and will replace glastpack as well as MRvcmt so
that everyone and everything is using the same build tool.
See
Riccardo's page for more complete information as well as a glimpse
of his other major project. [The latter could be heard objecting
to the meeting part-way through, but not as loudly as Richard's cat.]
- CMT, MRvcmt: (Toby) It used to be possible
to invoke MRvcmt with cmt path as argument so that multiple MRvcmts could
run concurrently, but this is now broken. (Riccardo) Leon found the
cause of this and it will be fixed soon. (David Chamont) Once we switch
over completely to the new CMT and are able to use cmt projects rather than
cmt paths, various problems people have had will disappear. Meanwhile,
he has a patch for a v16 problem involving CMT checking out a wrong
version under some circumstances.
Processing circus (Richard)
-
Some
runs needed as input to the Flight Software simulator
finished
on Friday.
- Reprocessing of DC2 sky is complete. Seth and Jim are looking at
it; no squawks so far. (Seth) No news is just that: no news.
He's not yet done with a careful examination to be sure that any differences
are solely due to the pulsar correction.
- Michael tagged a Beamtest release suitable for initial simulations.
Heather has written a ROOT macro
which will merge the merit tuple and an ancillary tuple file.
About 1 million events total will be simulated in a dozen or so configurations
requested by Benoit & Co. Would like to get this done by tomorrow,
giving people a couple weeks to look at it before the next Pisa meeting.
- GR9 validation ran into a snag because of particle name
changes in the new G4 version. Tracy has dealt with it by switching to
pdg id. There may still be an unpatched bug in OnboardFilter, v2. Short
term plan is to use v3 on Linux; there is no working Windows v3.
- Windows RM performance problems: (Toby)
RM Windows builds have been shut down for some time now. What's the
story? (Richard) A fair question. After some miscommunication,
3 days ago SCS finally started to defragment the V disk. The product
they use gives no indication of its progress. Hopefully this will
complete before too long and, if performance is still not adequate,
we can look for another cause. It has been very difficult to
get the attention of SCS Windows Support.
[Update from Navid a couple hours later: "Just to let everyone know: Windows RM is back up and catching up
with builds. The defrag on windows is done and there has been some
speed improvements on the drive but still not as much as we'd like.
We're still looking at other causes of the slowness."]
J. Bogart, Last Modified:
01-Jun-2010 15:48:09 -0700
previous
