2002 Scheduling Meeting
Attendance: Joanne Bogart, Toby Burnett, Richard Dubois, Marco Frailis,
Riccardo Giannitrapani, Hans Mayer-Hasselwander, Leon Rochester, Alex S, Tracy
Usher, Karl Young
Monte Carlo
- McParticle - design decision reached. Toby will implement the
class and Riccardo & Marco will fill the McParticles and put them on the
TDS.
- McPositionHitCol - hits are stored in a vector, sorted by id (NOTE:
the hit TKR strips within the vector are not necessarily sorted)
- detModel upgrades - 1 week
- GismoGenerator - 1 day Toby Upgrade to handle additional Monte Carlo
support is not in the critical path. In fact, this task has been given very
low priority.
- G4Generator - 1 week Riccardo and Marco **** what is involved here?
****
- instrument - 1 day Toby remove digitization and provide interface to
detModel for geometry information
- TkrRecon - Tracy, Leon, Toby remove GeometrySvc dependency and use the
geometry information through the instrument interface
- McIntegratingHit - 1 week Riccardo and Marco Input from the CAL group will
be obtained before implementation
In 2 weeks, Riccardo and Marco will complete their upgrades for the
MonteCarlo data and the G4Generator package.
Digitization
- CalDigi - Richard will contact Sacha concerning the status of CalDigi.
- TkrDigi - Bari's TKR digi algorithm will be added to the standard SAS CVS
repository. Riccardo and Marco will obtain the code and forward to Tracy and
Leon. Riccardo, Marco, Claudia will integrate the code and store it into the
CVS repository, with guidance from Tracy and Leon. Support for the BTEM and
BFEM data will be maintained. The mechanism for implementation needs to be
discussed further - whether or not to use Gaudi's subalgorithm
structure.
- AcdDigi Heather will implement the AcdDigi algorithms agreed to by Steve
and the ACD team at GSFC.
Reconstruction
- CalRecon - Richard will contact the CAL group
- TkrRecon - Transient Data Store (TDS) output needs to be settled. NOTE:
release in May will be non-optimized code 1 week next tag of TkrRecon using
the new GismoGenerator?? **** I never felt I got this straight ****
- AcdRecon - will wait for the next release of TkrRecon in May, when the
data objects stored on the TDS should be stable.
NTuple It was noted that the current ntuple structure is good enough for now.
**** I'm still surprised about this... I thought we were going to do an audit of
the ntuple contents **** The ntuple algorithms in the TkrRecon package will be
moved - such as the MC ntuple algorithms.
- Gaudi Upgrade to the new Gaudi v9 will take place once the modifications
for the Monte Carlo data are completed - 3 days Heather
- Traudl will be asked evaluate the Atlas' StoreGate.
- Persistent Data Store MC - new ROOT MC classes will be available the first
week of March - Heather
- Digi - need new ROOT Digi classes. Each subsystem has agreed to develop
their own. TKR will be implemented by Tracy and Leon CAL will be implemented
by Sacha ACD will be implemented by Heather These classes will be complete
in one month's time.
- Recon - new ROOT classes are waiting for design decisions by the
DataStructuresTF. For now, ROOT I/O will be handled by simple algorithms,
rather than a proper Persistency Service. This will be revisited in the late
spring.
RandomNumberGeneratorSvc - Toby will contact LHcB to see how they are
handling it.
ROOT upgrade to 3.02 We want to upgrade ROOT as we are writing the new ROOT
classes. Gloria has evaluated the new version for the existing EventDisplay and
it works just fine. We will have to coordinate tests for other portions of our
software.
FITS I/O interface for upcoming I&T support Will use the standard CFITSIO
library rather than the new ccfits, because ccfits does not yet compile on
windows. Heather will create an interface package and provide cfitsio in the
external library FTP area. Once the I&T group provides their FITS
specification, work will begin on a FITS->ROOT converter for the digi data.
Due August, 2002.
ROOT2IDL necessary for upcoming I&T support new version available August
2002
Calibration necessary for upcoming I&T support
- Need to evaluate MySQL versus PostGreSQL It appears that PostGreSQL
does not support windows well. Joanne will perform the evaluation and
contact Tony Waite for details concerning PostGreSQL.
- Setup DB on CentaurusA - Joanne will contact Pat Nolen
- Determine Meta Data
Setup tables for data
- Write interface to DB in C++ Madhup will be helping Joanne with the
implementation.
The subsystems are responsible for calibration algorithms. It was left
undecided what those routines would be written in: C++? ROOT? Currently our
calibration data is stored in ASCII files, in the future, will they be stored in
ROOT?
How does the interface to Gaudi work? Will the calibration data be put on the
TDS through the persistency service? Can we just use a home grown service that
will ingest the data and make it generally available? This was left for further
discussion.
Configuration Management - We will hold steady with gcc 2.91 We discussed using
Gaudi as an external library need to provide mechanism to run debugger when
necessary, by having both DEBUG and release versions. New version of VCMT
developed by Karl and Alex in the next 2-3 months, using SOPE + PERL.
ReleaseManager, the server version, available in 2 weeks.
Upgrade to use the standard units: mm and MeV Flight.xml uses the standard
units (mm and MeV). It is unclear how and when other code will be upgraded. It
was noted that the origin has been moved to the appropriate location, between
the TKR and CAL.
Documentation - semi-monthly code documentation review and code walkthroughs to
begin in April. Web doc review to begin immediately. Janani is currently going
over our existing how-to pages. Her current status will be queried early next
week.
**** Things we didn't address: ***** when/how will we upgrade to Geant 4.4?
H.Kelly Last Modified: 2002-03-06 09:41:59 -0800