The results of the comparison projects presented in the previous day’s workshop sessions were discussed. The presence both of good agreement in some comparisons and differences in other comparisons was noted.
The wish to see the active participation of more people in the physics comparisons was expressed. Also the need to quantify the level of confidence provided by chosen experimental comparisons was noted.
A summary of common differences seen by comparison projects was made by Michel Maire. Then he proposed a plan of work of the EM (standard) working group to verify the implementation of key processes, dealing with four aspects: stepping, energy loss fluctuations, multiple scattering and the incorporation of geometrical information in suppressing secondaries (“energy loss plus”).
Conclusions:
1. An effort to obtain updated timetables from the comparison projects for their progress was requested.
2. Plans were presented the EM (standard) working group to review the implementation of processes in four stages as a response to results reported by the comparison projects.
Atlas:
Atlas requested support for shared libraries.
LHCb:
LHCb requested
HARP:
HARP requested that measures be adopted to minimize the initialisation time of Geant4 program runs. Currently for the case of full detectors that contain on the order of a hundred materials and ten physics processes the initialisation time obstructs the development turnaround time of a simulation.
Different aspects of this issue were discussed, and the TSB agreed to create a plan for reducing the initialisation time of processes. Options mentioned included storing physics tables with high calculation cost and the possibility of reducing the re-computation of physics tables.
ESA:
ESA requested that a record be kept of people who download Geant4, the platforms that they are utilising and their area of application.
In response the TSB requested that a detailed proposal be submitted of the information to be collected, whether it should be compulsory or not, and other details.
BaBar:
BaBar requested:
1. The ability to choose processes based on a volume.
2. The ability to define cuts based on volume.
3. Multiple scattering: -- details not noted …. BaBar representative please provide them.
4. Assistance in choosing an appropriate set of hadronic processes.
In response TSB members noted that a method to choose processes depending on a volume was demonstrated by a class in the Geant4 source code: G4IonisationByLogicalVolume.
In contrast the ability to define cuts based on volume was a potential feature purposely suppressed by design in Geant4 – in order to assist in obtaining coherent physics results.
The requirement of several parties to establish a list of Actions for the TSB was agreed.
The proposal of Category Coordinators to release Geant4 3.0 on December 15th after a six week release period was agreed.
The schedule for the release of the documentation for this release was discussed. It was agreed to release the installation guide at the same time as the source code and to release the remaining documentation on December 22nd.
BaBar requested the support of Solaris 7 and the C++ compiler CC 5.1 .
This compiler is currently being tested as part of Solaris certification at CERN, and results will be made known at a future meeting.
The need to notify users of plans to stop support of some RISC platforms at a date to be specified (and no later than December 2002) was brought forth. It was agreed to survey Geant4 partners to accertain their needs and plan the stopping of support of HP and DEC(alpha) platforms.
The potential to assist in the use of Geant4 in new application fields was mentioned.
However the need to create a workplan for the furthering of such activities was noted.
The ability of Geant4 partners to utilise Geant4 in new areas of application was recognised.
The urgent requirement of a disclaimer of liability to accompany the release of Geant4 code was noted.
The IGD group brought forth four new candidate collaborators.
Walter Rolando requested to join the low energy EM group.
Unless noted, default date for actions will be the next TSB meeting.