Minutes of the Geant4 Technical Steering Board Meeting,
In person: John Allison, John Apostolakis (chairman), Makoto Asai, Guy
Barrand, Frank Behner, Gabriele Cosmo, Gunter Folger, Juergen Knobloch, Hisaya
Kurashige, Michel Maire, Youhei Morita, Maria Grazia Pia, Florence Ranjard, Takashi
Sasaki, Marc Verderi, Johannes-Peter Wellisch
By video: Katsuya Amako, Peter Gumplinger
By telephone: Peter Truscott, Hajime Yoshida
Reports from Working GroupsGeometry
(G. Cosmo): a few minor fixes and a protection added in trasportation were
included in last reference tag. A new STEP interpreter will be available shortly;
feedback will be appreciated. P. Truscott volunteered to try it and provide feedback
to the Geometry group, or at least provide example STEP files [that have been
found to be problematic for the current STEP reader].
The issue reported at
CHEP, concerning the implementation of the geometry of Atlas calorimeter, is being
investigated, but it is anyway evaluated as at least an order of magnitude lower
than what reported by the user.
Run and Event (M. Asai): the EventManager
becomes a singleton. Two bugs have been fixed.
Generic processes (M.
Verderi): a proposal for development will be represented. A couple of bugs have
Standard electromagnetic physics (M. Maire): the group
focuses on testing; a photoabsorbtion model is under development.
physics (J.-P. Wellisch): the plan for the Pythia7 interface is in progress.
The Kinetic Model approaches release. Fixes for strings in file names and for
unit problems have been introduced.
Intercoms (M. Asai): it had a fix.
Visualisation (J. Allison on behalf of S. Tanaka): new commands have
been implemented. A prototype Ray Tracer, developed by M. Asai, is available.
Persistency (Y. Morita): a few changes for boolean solids and persistent
hits have been released. An Extended Example is available. Persistency has been
included in system testing.
Testing (J. Allison): extended examples
are part of the system testing; there have been some minor improvements in scripts.
A web-based control and monitoring system based on Bugzilla is in progress. The
possibility of moving to Remedy as a tool for problem reporting was discussed.
Software management (G. Folger): an ISO-compliant version of Geant4
is available; as not all compilers are fully ISO C++ compliant, i.e. no support
for namespaces and/or native STL, a non-ISO version will still be maintained for
some time too. Coding guidelines, amended by G. Cosmo for ISO C++ compliance,
are posted on the web.
Documentation (K. Amako): a new set of documentation
has been released with Geant4 1.0; a web page reporting documentation changes
has been available too. The issue of modification to the documentation, in the
case any fixes are required after it has been posted on the web, was raised; the
practical way how to handle it will be proposed by K. Amako to the TSB mailing
Low energy electromagnetic physics (M.G. Pia): the domain is
under dynamic development; currently the activity is focused on testing and validation.
Processes for low energy hadrons and ions have been released; new physics models
are under development to further extend the energy range coverage for antiprotons.
A paper has been published on Geant4 low energy electromagnetic processes for
ions. A design and code review is in progress. The activity is being reorganised
with the creation of the Low energy electromagnetic physics Working Group.
Apostolakis will circulate a document with a refined definition of the milestones
soon. He requested proposals of intents for joint projects with experiments within
one week, to be reviewed in two weeks' time.
J. Apostolakis requested a draft
of proposal on the Geant4 training kit from the user communities in two weeks'
time. Training material is being developed for the CERN School of Computing by
M. Asai, G. Cosmo and M. Verderi. M.G. Pia will give lectures on Geant4 at the
Detectors School in Italy in February 2000.
A discussion on software process
improvement was encouraged and a draft plan was requested by J. Apostolakis in
Working group milestones will be discussed on the TSB mailing list
and will be reviewed at next TSB meeting.
was noted that the regular release of monthly reference tags to the member user
communities will help to incorporate feedback from experiments as a kind of pre-release
Some aspects of the Problem Report System were discussed. After some
discussion it was agreed that (a) it should offer the user the possibility of
selecting multiple platforms and versions, (b) that it should be easier to see
the list of users and developers known to the system, (c) that email addresses
not known to the system should be allowed in the Cc: box and (d) that the restriction
on the visibility of reports would be lifted.
A FAQ web page will be available;
G. Folger will assemble it, with content provided by others.
modules developed as part of the ESA contract will be delivered in March.
Pia requested the availability of a design repository, to be kept updated. She
also requested to maintain and regularly update the Geant4 URD. Both these requests
will be part of the software process improvement plan.
TSB members should
help to create a publications plan.
The CERN/IT/API group will create a poster
to document the Geant4 activity in the IT building; other communities expressed
interest in it.
Subgroups for the preparation of the TSB meetings agenda,
for web pages and for software process improvement were proposed. J. Allison and
J. Apostolakis resigned from the existing web working party.
is a new member of the TSB, replacing David Jacobs.
Stéphane Chauvie is a
new member of the Italian Geant4 group.
Next meeting and
Annual WorkshopProposed dates for next TSB meeting are 29/3 or 4/4/2000.
It was proposed to hold the next Annual Workshop on 11-15/9/2000 in Paris.
M.G. Pia commented that these dates could be problematic for the Italian Geant4
collaborators, because of expected overlap with the annual INFN budget meetings
for year 2001; she suggested to hold it later. An overlap with the CERN School
of Computing (Geant4 lectures on 18-22/9) should also be avoided. M. Maire and
J.-P. Wellisch noted that they cannot be available during October; there are the
Monte Carlo conference, and the calorimeter conference. The dates will be further
Pia, 1 March 2000