Minutes of the Geant4 Technical Steering Board Meeting

16 June 2000

Attendance:

CERN: John Allison, K.Amako, John Apostolakis (chairman), Guy Barrand, Gabriele Cosmo, Gunter Folger, Juergen Knobloch, Hisaya Kurasige, Ramon Nartallo, Maria Grazia Pia, Florence Ranjard, Matthias Schroeder.
SLAC: William Lockman. (morning session)
TRIUMF: Fred Jones (morning session).
KEK: Makoto Asai, Youhei Morita, Takashi Sasaki,
By telephone: Irena Gudowska, Hajime Yoshida

Apologies for absence: P.Gumplinger, M.Maire, M.Verderi, J.P.Wellish

Minutes: F.Ranjard - 7 July 2000 pm

Managerial issues

New Geant4 members:
- Karolinska Institutet,
- HARP
New TSB members:
- Atlas: Katsuya Amako
- KEK: Tahashi Sasaki
- Karolinska: Irena Gudowska
- HARP: Juan Jose Gomez Cadenas
- CMS: Matthias Schroeder (deputy)
- ESA: Ramon Nartallo (deputy)
Changes in STT group:
- new coordinator: Serguei Sadilov
- new member in CERN/IT : Steve O'Neale.
Changes in Hadronic process group:
- one new member in Frankfurt group.
Memberships in Low Energy Electromagnetic Physics group
- new members in IGD group: Stefano Agostinelli, Stefania Garelli, Giovanni Santin.
- Some concerns were expressed about affiliation of already existing contributors to newly collaborating parties, including cases where this would affect existing work responsibilities. These concerns will be addressed to the Collaboration Board.
Review of manpower within Working Groups:
- J.Apostolakis asks the WG coordinators to reporet within one week on the manpower used
- in the past 6 months (names and time).
- The WG coordinators should prepare their manpower requirements for the next TSB.
Absences during Summer time:
- John Apostolakis : 10 July - 10 September
  - deputized by Makoto Asai
- Peter Gumplinger : 1 June - 1 September
  - deputized by Fred Jones
Geant4 workshop in Paris:
- 16 -20(21) October
- A Web site will be opened soon
next TSB meeting:
- possible dates: September 15, 22, 29, October 6, 18, 19 (at Workshop)
- following dates are not possible for KEK (15/9), LHCb (29/9), CMS (6/10)
  - decision will be taken by the spokesperson deputy between September 22 and October during the workshop.
Minutes of last TSB meeting:
- Minutes of the meeting held on 27 March 2000 were approved.
Report from CB meeting held on 13 April 2000:
- The comparison projects proposed by ATLAS, BaBar, CMS and ESA were discussed.
- There was no objection for encouragement on these proposals. It was a consensus that
  these proposals have to be studied more on detail in the manpower, complementarity, band schedule for the refined plan.
- Low Energy EM working group was formally approved.
- new full membership of Karolinska Institute and HARP collaboration.
- Discussion forums to facilitate exchange between users have been accepted and will be implemented using Hypernews.
Notice of Application of a new group:
- Proposal of a UK group to Join Geant4 presented by J.Allison.
- This proposal has to be put forward to the Particles Physics and Astronomy Research Council in UK and to the Geant4 Collaboration board for approval. The group will be composed of: J.Allison (1 CU), S.O'Neale (0.4 CU),
  G.Blair (0.2 CU), G.Hughes (0.1 CU), P.Kyberd (0.2 CU), G.Patrick (0.1 CU).
  - The TSB is favorable to the application of this group.

Following up of 2000 milestones, CB and TSB decisions

Comparison projects

BaBar: W.Lockman presented a revised version of the proposal
- Focus on how well G4 simulates the data obtained from BaBar detector tracking system
- (SVT) and a 40-layer drift chamber (DCH): delta ray distributions, dE/dx and multiple scatterings.
- time scale according to experiments'needs:
  - start as soon as possible, duration ~ 1 year, ~ 2 FTE.
  - requested expertise from Geant4: an expert in e.m. physics processes and if possible in geometry description.
  - development of new models, if necessary, should be discussed within the working group.
  - Geant4 manpower: ?
ATLAS: K.Amako presented the outlines of the project
- compare G4 simulated data with test beam data in:
  - EM barrel calorimeter
  - Hadronic endcap with forward calorimeter
  - Hadronic barrel calorimeter
- to test electromagnetic and Hadronic shower processes.
- test beam data exist already and comparison has already started.
- Atlas manpower (very preliminary)
  - EM barrel: 4.5 FTE
  - Hadronic endcap with forward calorimeter: 0.6-1.0 FTE
  - Hadronic barrel: ~ 2 persons?
- Kickoff meeting : end of June
  - It is confirmed that the comparison project does not mean any exclusion of the potential support
    of Geant4 collaboration to any other problem which might be found in the collaborating experiment.
- - Geant4 manpower: ?

CMS: M.Schroeder presented a revised version of the proposal
- validation of Si test beam data: energy deposition.
  - mean signal is well defined by Geant3 but not the shape.
- data taken in 97 and 99
- beam 8GeV and 100MeV.
- time scale : ~ 6 months starting in September.
  - The TSB would like to get more detailed on people involved.

Training milestone

M.Asai is rewriting his single day training session.
A training web page is available from the INFN Geant4 web site including the training material developed by
M.G.Pia and contributions by other Collaborators.
It was agreed to adopt the User Requirements Document presented at last TSB meeting by ESA and INFN
collaborators as the initial ERD for the Training Milestone. M.G.Pia agreed to do the minor editing required
to gibe the document a more general layout. A spiral process of updated of the URD and Training Kit developments
can be forseen.
- time scale: ~1 month.
Commitment from the Working Groups and their Coordinators to contribute to the Training Kit is needed.
M.G.Pia invited the training party to articulate the preparation of the training material as a modular kit, out
of which the trainers can assemble training sessions suitable for various applications.
- A first draft should be circulated mid September.
Proposed workplan:
- Mid July : first skeleton for each kit
- Mid September : first draft for each kit
- End September : feedback from category coordinators
- October 17 : presentation and review at workshop and feedback from collaboration.
- November 10: final version of text.
- November 20: final version of exercise tested with candidate tag of December release..

Software process milestone

G.Cosmo presented a workplan & timescale of the Software process activity:

Q/A and Optimization activity.
- the creation of the Q/A & Optimization team proposed last TSB should be done as soon as possible.
  - first application : release 2.0 (30 June 2000)
- get more resources from user groups as soon as possible.
- The team should implement proper filters to enforce coding conventions
  - dead line: 30 September 2000.
- make all the tools available to developers.
Analysis & Design software cycle
- Review the current category diagram.
- Generate a questionnaire based on ISO-15504 (references: CERN Project Support Team and
  past SPICE assessment) and distribute to Category Coordinators.
- Collect results of the questionnaire and integrate assessment's results with Software Process improvement proposal in a ISO-15504 document.
- Define maintenance and update procedure of documents/diagrams.
- Update documents/diagrams.
- Create a CVS repository with restricted access to G4 Category Coordinators for collecting
- architectural designs - as soon as possible.
Testing
- Review currently dedicated resources available for testing - as soon as possible.
- Recruit new manpower and promote a training activity.
- Establish clear responsibilities for maintenance and integration of system tests and examples.
- Review and improve code quality of public examples.
- Review and document current system tests.
- Use regression tests
- Use statistical tests
- Improve/implement automation:
  - Adopt Bonsai (May 2000), LXR browser and Tinderbox (September 2000)

Open discussion forum

Forum will be segmented into categories with no commitment to answer questions.
Forum will be implemented with Hypernews.
SLAC will be the host and the maintainer.
- - installation starts these days.

Geant4 paper

A paper should be written by the end of the year.
J.Allison could be the editor.
- - Get a skeleton in the next month . People involved: J.Allison, J.Apostolakis, M.Asai, J.Knobloch.

Geant4 Community life

current release status report

STT report (J.Allison)

testing is done 3 groups with S.O'Neale efficient help
- group1 = low category is OK
- group2 = management and E.M. physics
- group3 = Hadronic physics, interfaces and visualization
Hadronic physics is being tested in parallel.
- - Final release end of June.
Release manager report (G.Cosmo)
- problems with complex numbers are solved.
- memory leakage (insure++) in Hadronic physics is being solved.
- unfortunately the test does not cover all processes.
  - review the existing tests for memory leak and extend the coverage.

reports from working groups

Software Management: policy for reference tags
- a tag should handle one development or one bug fix at a time.
- a tag should be well documented.
- category coordinators should announce Problem Reports solved.
- user groups should give feedback quickly after the monthly release which solved their Problem Reports.
- deputy spokesperson to collect user group reports.
System Testing (J.Allison):
- Bonsai is now installed which allows to query the CVS repository tree.
Low Energy Electromagnetic Physics recent developments (M.G.Pia)
- The developments take place according to the WG Milestones for year 2000 presented in March.
- A joint project with users at the Italian National Institute for Cancer Research is in
- progress for the comparison of Low Energy photon processes with experimental data.
- Excellent results are already available on the calculation of photon attenuation coefficients.
Visualization Mini Workshop status report (S.Tanaka)
- Following requirements were recognized:
  - requirement for Geant4 to fit into existing Interactive Frameworks (Momo,
  - Explorer, AVS, BaBar, Gaudi, OPACS, JAS, WIRED, ROOT)
  - Clear need to update Geant4 policy on histogramming
  - extend the "Delegation" principle to G4Trajectory, G4Axes and G4Text.
  - display geometry hierarchy and attach interactive functionality to the nodes (DTREE).
  - dump volume and/or trajectory properties when clicking on a view.
  - a way of selecting on the G4 side the objects passed to the viewers or framework.
- Several actions are still pending.
- Documentation and FAQs should be improved for June release.

reports from user groups

IST Genova

M.G.Pia has shown very interesting Geant4 results from a brachytherapy application.

ATLAS

presents the following requirement: support of HepMC event generator interface. Several questions should be answered:
- Is HepMC the unique interface?
- What about StdHep++?
- Which interface is used by Pythia7?
A CLHEP workshop end of June at Fermilab will clarify the situation.
- - Event category coordinator will collect the information.
  - Geant4 will provide an example in "extended" category for demonstrating how to interface with HepMC.
  - Having new dedicated Geant4 class(es) in Event category is suspended till the above questions are answered.
Guy Barrand and al (P.Binko/LHCb, T.Johnson/SLAC, A.Pfeiffer/CERN)

present the following proposal: implement a common interface to various analysis tools.
- - It is agreed to create an "analysis" directory (not yet a "category") to be able to
  - implement examples which use various analysis packages: G4analysismanager, JAS, OpenScientist.

Content of next major release

Platforms and Compilers:

In view of the December major release it is necessary to set a list of platforms and

compilers used by the Geant4 user community.
platforms and compilers used by Geant4 should be in phase with other package choices (LHC++, Objectivity,etc...)
RISC platforms are being phased out.
- The TSB users representatives should set the list of platforms and compilers used by
- their groups. The final choice will be made at the next TSB meeting.

CERN Accelerating science