Minutes of the Geant4 Technical Steering Board Meeting
22 May 2001
Attendance:
CERN: Katsuya Amako, John Apostolakis (chairman), Pedro Arce, Gabriele
Cosmo, Gunter Folger, Vladimir Ivantchenko, Maria Grazia Pia, Florence
Ranjard, Sergei Sadilov, Matthias Schroeder, Johannes Peter Wellisch
SLAC: Makoto Asai, William Lockman, Dennis Wright
Ecole Politechnique : Marc Verderi
KEK: Youhei Morita, Takashi Sasaki
Manchester : John Allison
Apologies: Peter Gumplinger, Jurgen Knobloch, Peter Truscott, Petteri
Nieminen.
Minutes: F.Ranjard - 22 May 2001
Regular items
-
Minutes of last meetings
-
J.Apostolakis will circulate this week the updated version of the minutes
of the March meeting - all comments are due by May 31st.
-
He will circulate a finalized version of the minutes of the January meeting by the first week of June. One week after their release they will be
final.
-
Review of pending actions
-
the action list will be kept separate from the minutes and appended to
the minutes.
-
actions on the action list should contain the opening date and a foreseen
closing date
-
look at the action list appended to the minutes
-
Report from meetings
-
CB meeting held on 25 April 2001
-
memberships of Barcelona and Helsinki are accepted.
-
a warranty
draft is being circulated: if no opposition it will be considered to be
accepted.
-
the reference to the warranty should be included in all files at the next
scheduled public release.
-
Software process goals and objectives for 2001 were well accepted.
-
preparation of the review
-
minutes
of the Category Coordinators Meeting held on 23 April 2001 are available
-
category coordinators are asked to provide a list of authors for the Geant4
paper.
-
Request from experiments
-
CMS
-
requires implementation of reflections in geometry
-
needs a tool to debug geometry implementation
-
action M.G.Pia: foresee a general presentation of various tools for geometry
checking in a parallel
session of the workshop.
-
BaBar
-
requires implementation of an interface to allow the cutoffs to be different for
different large-scale regions of a detector (one way of implementing this would
be to specify the cutoffs for each material.)
-
action K.Amako, D.Wright, J.Apostolakis: document the Atlas/ BaBar user requirement
for different production thresholds in geometry regions and submit it
to the TSB (by July 1st)..
-
action H-P Wellish: provide a requirement template and submit it to the
TSB.
-
LHCb is working on a build of G4 dynamic libraries on Windows. This implies
to add few lines in the header of several include files. When the work
will be finished and proved to be fully operational a request will be submitted
to the TSB to commit these changes to the CVS repository.
-
before accepting this request the TSB would like to see the changes foreseen
because there is no machine dependent code in Geant4.
-
ATLAS requests the improvement of the performance for complex volume
parameterizations.
(Note: Requirements should be presented to the TSB using the TSB
mailing list, not through G4 HyperNews.)
Geant4 Hypernews Public Forum
-
site is ready and should me made public
-
action: add a description of the purpose of the forum in its first page,
explaining that the forum is meant for the exchange of G4 expertise between
users, and that the collaboration will respond to questions on a best effort
basis. ( M.G. Pia + J. Apostolakis to draft by May 31st and send to the
TSB).
-
action: create a link to it in the G4 home page (dead line : 5 June)
Preparation for review
-
a draft plan for the
review week (18-22 June) and its timetable
are available
-
The TSB regrets that no time is foreseen for presentation of user applications
not included in a comparison project.
-
action: J.Apostolakis will make sure that some time is available for such
presentations.
-
J.Apostolakis asks every experiment to be ready at least for a 15 minutes presentation.
-
presentations should be available for comments a week before the review.
-
some members of the TSB feel that the sharing of time between e.m. physics
and hadronic physics does not reflect the weight of the code related
to these areas.
-
"Part II : Focus on lifecycle processes" should be coordinated by G.Cosmo
with emphasis on:
-
how design process is done
-
validation
-
software construction and maintenance
Preparation for Workshop
-
a Web site and
proposed Program are available: TSB members are asked to send their comments
to the program committee: M.G.Pia, M. Schroder, H. Yoshida.
Working group milestones
-
For a full
list look at the minutes of the Category Coordinators meeting 23 April
2001 .
Software Process
-
G.Cosmo presented the
action list:
-
progress should be made fast, before the review.
-
design diagrams for each category are required.
-
design documentation should be updated as the same time as code releases.
-
M.G.Pia asks each category coordinator to come to the workshop with a written
list of few user requirements to be used in a tutorial parallel session.
-
Action: G.Cosmo, MG Pia, HP Wellisch, J Apostolakis to propose workshop project
on Software Process. (Time: 1st June)
-
Current status and actions:
- the following category architecture diagrams are missing, to be committed:
-
Tracking, track
-
Events, Run, Digits+Hits
-
Intercom, Interfaces
-
Visualization
-
Analysis
-
Design of Global, Particles and Materials are mixed
-
G.Cosmo will split the files in 2 parts.
-
Design diagrams of Geometry, Process, Persistency need updates.
-
Revised
Release Policy
-
a version number is of the type xx.yy (major xx, minor yy)
-
major release only if user-visible code is not backward compatible or if major
new functionality is added, and only at scheduled release dates fixed by TSB (eg
it could happen in 2001 only at scheduled dates of public releases of June 29th
or December 14th).
-
minor release must not break user code, but can contain new features.
-
if the July release remains backward compatible it will be a minor release.
Architect Subgroup of Category Coordinators
The tasks of this subgroup will be:
-
maintains URDs and Design documentation.
-
reviews significant design changes proposed by category coordinators
-
recommends design changes to the TSB.
Publication
-
action : next TSB should fix a schedule for the Geant4 paper.