Present: Joanne Bogart, Toby Burnett, Richard Dubois, Heather
Kelly (mute), Michael Kuss, Leon Rochester,
Alex Schlessinger, Tracy Usher, Karl Young
-
System Tests:
(Karl) They can now be run "by hand" for GlastRelease.
Still a bit of work to do in scripts (e.g., getting all required
environment variables set correctly) before they can be run
automatically.
-
GlastRelease/Nightly builds:
(Alex, Toby, et. al.)
- Basically in same state as last Thursday: some tests are still
failing.
- Would be helpful to be able to find out easily what has changed
from one build to another.
- Problem with current build of v0r3: reports tests 0/0.
Richard inadvertantly clobbered the old build when he tried to make
a private build. Alex has since changed the protection on the
release area to prevent such incidents in the future.
- GlastRelease and all external libraries except G4 (will soon get
binaries Heather has built in correct environment) has now been
built with Alex's build of the gcc 2.95.3 compiler so that all they
all depend on the version of system libraries. Will soon incorporate
Heather's G4 binaries.
- Using ROOT 3.04 on Linux can get error "problem with X
handler", after which many things don't work. [This was quickly
responded to by the Root folks - they added capability for hardware
assist, but not all X emulators handle it. SLAC's X-Win32 Windows
X-server appears to fall into that category, and a RootSys system.rc
parameter has been modified in the SLAC Root install of 3.04.02 to fix
this. Ed]
- When do we declare a new release? The only GlastRelease release we
have is now quite old. Haven't been at a point of stability where a
release might make sense for some time. Partly this is a result of
the change to new versions of G4 and ROOT, but current problems are
unrelated to those upgrades.
- (Leon) Having difficulty assembling a compatible collection of
package versions for the work he is doing.
Calibration infrastructure:
(Joanne)
-
Working on calorimeter. Have written a draft XML format (a
temporary measure; ROOT is preferred format for per crystal face/range
data). Need to complete TCDS class description(s). Current design for
these involve templates; not yet sure whether this will play with Gaudi
converter mechanism or with ROOT.
-
During discussion on Friday Toby (again) raised the possibility of
using calibration infrastructure or a near clone to handle parameter set
data. I plan to look into this after calorimeter support is in better
shape.
-
Other
activities.
-
Iterative recon: (Tracy) Will go
into Recon package. Depends on Bill Atwood's improved energy
calculation, which in turn uses vertexing information if
available. Hence the new scheme, unlike the old,
does track fitting and vertexing as well as pattern recognition on the
first TkrRecon pass. The algorithm driving the iteration is also in
TkrRecon. [Many thanks to Tracy for explaining this in detail
post-meeting. ed.]
-
Calorimeter recon: (Richard) Has
Re-Tooled CalRecon, so it now has Gaudi Tools for clustering and energy
leakage. Next
is to get it working with the Recon package modifications Leon is
preparing with Bill Atwood's energy correction code.
-
Trigger rates: (Richard for
Steve) trigger rate is lower in Gleam than in pdrApp, in particular rate
for albedo gammas is about half of old rate. Should we think about
reviving BTEM data for Gleam? [ouch! ed.] One possibly significant
difference is using G4 rather than gismo. (Leon) What about the .7 range
cutoff? Could try varying this number to see if it has an
effect.
-
News and notes: (Richard)
-
All CHEP03 papers but
one have been accepted. The missing one (on DPF prototype) couldn't be
prepared on time anyway since the work it would describe has been pushed by
other more pressing matters. [The TkrRecon paper was accepted as a posted,
and will be withdrawn. ed]
-
Tracy will be gone next week.
-
Action items: A gremlin posted
the page to remind us to review them. See updates below.
J. Bogart Last Modified:
04-Aug-2004 15:39:16 -0700