Minutes of TSB meeting #15

20 November 2001 - 9:00-13:00

Attendance: 

CERN: John Apostolakis (chairman), Pedro Arce, Gabriele Cosmo, Gunter Folger, Vladimir Ivanchenko, Florence Ranjard, Albert De Roeck, Sergey Sadilov  

Ecole Politechnique: Guy Barrand, Marc Verderi

ESA: Petteri Nieminen, Peter Truscott

KEK: Youhei Morita, Takashi Sasaki

Kobe: Hisaya Kurashige

SLAC: Makoto Asai, Dennis Wright

TRIUMF: Peter Gumplinger

Excuses: J. Allison, K. Amaco, I. Gudowska, J. Knobloch, B. Lockman, M. Maire, M.G. Pia, N. Starkov, S. Tanaka, H. Yoshida, H.P. Wellisch

Minutes: V.Ivanchenko, 6 December 2001 first draft, 11 January 2002 v0.3

Regular Items

Development Objectives and Release 4.0

• Summary from Working Group reports at Category Coordinators meeting:  
  • Status of WG development for release 4.0 was reviewed.
  • Milestone of 2001 are achieved with few exceptions, details are described below.
  • First step to design regression suite will be one example developed Low Energy EM group.
  • Plans for an analysis category: examples/tests making use of histograms/ntuples will adopt AIDA interfaces directly in the code. The code in the analysis directory is obsolete, as functionality is in AIDA.
• Open issues:  
  • The status of hadronic physics: it was pointed out that there are many requests from users to provide detailed documentation about G4 hadronic processes. This subject was discussed during Genova G4 Workshop. In the release 4.0 hadronic package will be upgraded significantly, CHIPS model will be delivered, several new system tests focus on hadronic physics will be used. One of 2002 milestones is expected to be to provide a set of examples of PhysicsLists with hadronic physics for different use cases.
  • New action required to add automatic notification to the problem report system.
• Underlying libraries/frameworks
  • CLHEP 1.7 is delivered and the release 4.0 will be maintained on top of it. CMS report a problem with usage of one CLHEP class, investigation is required by their part. 
  • AIDA  2.2 will be delivered next week. The new release will include n-tuple implementation, that will be used, at least, in one G4 advance examples.
  • Visualisation working group have done a review of the set of environment variables required for visualisation. The number is decreased to minimum in 4.0.       
• Release time schedule is published  
  • New item: Documentation of new features is mandatory. To be checked on December,1.
  • The release day December, 14.
  • Currently the main problem is time penalty indicated in some tests. 

Regular items

• Reports from experiments/institutes/user-representatives
  • F.Ranjard LHCb:
    • G4 activities are concentrated on two LHCb subsystems: ECAL and RICH.
    • G4 results for ECAL are satisfactory and CPU time is close to that of G3.
    • A G4 server are available and installation of geant4.3.2.ref07 is in progress.
    • Old requirement from LHCb: provide flexible G4Exception to allow application to customise user actions and not kill the job. Already implemented and will be included in release 4.0.
    • There is a problem with usage of B-meson lifetime, investigation is required by their part.
    • There is a request to move Boolean processor from the Geometry category to CLHEP, because it is used in LHCb visualisation independently. To be discussed in the frame of CLHEP. Evgueni Tcherniiaev should be involved.
    • Problem report #305 will be reassigned to right person in Bugzilla and required investigation. 
  • P. Arce CMS:
    • Full simulation of CMS is available now.
    • There is a problem in G4BREPPolycone, which is under investigation.
    • Performance comparison: G4 -1111 tracks/s, G3 - 1153 tracks/s.
    • At small angles G4 performance decreases by order of magnitude probably due to of slow transportation in magnetic field and low energy particle transport in vacuum.
  • P. Nieminen ESA:
    • Two new tools ESA-sponsored will be delivered in January: GEMAT (Geant4-based Microdosimetry Analysis Tool) and MULASSIS (Multi-Layered Shielding Simulation Software).
    • There are plans for 2002 to re-engineer General Particle Source and to implement internal conversion for deexcitation of excite nuclei.
  • P. Truscott ESA:
    The absence of the Intranuclear Cascade code (the re-engineering of HETC) presented a significant problem to MULASSIS and GEMAT usage, which requires an accurate treatment of (0.1 - 5.0) GeV hadron-nuclear interactions.
  • D. Wright BaBar:
    • 100 Million events have been produced in G4 Bogus production run.
    • SLAC has agreed to assist TRIUMF in maintenance of G4 parametrised models.
    • New requirement: to provide a possibility to converge BgsTransportation and G4Transportation. Design and analysis are required.
• G. Cosmo reported of a new requirement collected by M.G. Pia and coming from astrophysics community, for providing simulation of high energy cosmic rays interactions.
• The solution of the problem will be achieved on base of model approach for electromagnetic physics. The discussion of design of the model approach is in progress in EM working group and low energy EM group. It is assumed to be a milestone for 2002.

Documentation review 

• Introduction is rewritten. FAQ of use cases are updated.
• Installation guide is 70% done.
• Application Developer Manual is well advanced.
• Toolkit Developer Manual updated for about 25% of the requested changes only.
• Physics Reference Manual for EM physics is in a good shape, hadronic physics description is incomplete.
• The range of applicability of physics processes requires explicit attention.
• A clear contact person for each physics process or model is desirable.

Software Process 

• Architecture group was nominated: M. Asai (chairmen),  G. Cosmo, H. Kurashige,  H.P. Wellisch. 
• Process for tracking new requirements: automatic system for submission of user requirements is in progress. 
•  S. Sadilov: progress in automation of system testing connected with modification of Bonsai selection for testing, TinderBox is installed, and LXR will be installed next week.
• G4 design diagrams for a number of categories were obtained by M.G. Pia using inverse engineering with the Rose tool. For the next TSB the status of Rose access in different centers should be presented. 

Publications

• M. Asai supposed to provide a schedule for the G4 architecture paper.
• G. Cosmo will be responsible for publication on G4 software process.

Announcements

• M.Asai inform about Geant4 Users Workshop held at SLAC from 18.02.02 to 22.02.02, which will include G4 tutorials and original presentations about G4 applications.

Milestones 2001

• The detailed description of development and new functionality is presented in the summary from Working Group reports at the Category Coordinators meeting.

• Major part of milestones for 2001 are achieved and implementation will be available in 4.0 release. 

• The general G4 paper is in progress. 

• The documentation has been upgraded. The Training kit was exercised in connection with several tutorials. 

• The comparison projects were in progress and publications on these projects are assumed to be a milestone for 2002. 

• Some elements for the regression suite have been implemented by the Low Energy EM Group. 

Milestones for 2002

• Process for creating milestones of 2002 will be based on the requests of experiments and user groups.
• Experiments and user groups are requested to send proposed requirements now.
• First proposals of milestones for 2002 will be drafted to the next TSB meeting which is to be scheduled for January.
• In addition a system for collecting new user requirements will be put in place.

Report from Event-biasing mini-workshop

• Use cases
• 'Leading' biasing techniques
• Prototypes for 'easy-to-use' biasing.

AOB

• Next meeting January

Status of pending actions

Action 9.6

An example to use HepMC as an external package will be provided in 4.0.

Action 11.19

G. Folger need help to do statistical analysis.

Action 14.4

First half is done, second is opened.

Action 14.6

S. Sadilov is working on.

Action 14.9

To be ready for the next TSB.

Action 14.10

S. Sadilov is working on.

New pending actions

Action 15.1

D. Wright to organise the second stage of global review of Geant4 documentation - detailed review (plan and schedule by January 2002).

Action 15.2

G. Cosmo to report at next meeting about progress in implementing removal of inline static methods from the code, in view of facilitating porting on W2000 for of shared libraries.

Action 15.3

J. Apostolakis to identify one testbeam example for which analysis will done using AIDA interfaces.

Action 15.4

J. Apostolakis to send to CB chair a request to finalise the review report.

Action 15.5

To all responsibles in the main G4 labs, report on the availability Rational-Rose and Together CASE toolkits in the respective centers.

Action 15.6

M. Asai to provide flexible G4Exception, corresponding example, and documentation.

Action 15.7

G. Folger to add automatic reminder notification to the problem reporting system.