Attendance
- CERN : Katsuya Amako, John Apostolakis (chair), Guy Barrand, Gabriele Cosmo, Gunter Folger, Michel Maire, Youhei Morita, Maria Grazia Pia, Florence Ranjard, Matthias Schroder, Hans-Peter Wellisch (until 10:45).
- KEK : Takashi Sasaki, Makoto Asai.
- TRIUMF : Peter Gumplinger (first session)
- Manchester: John Allison.
- Telephone (first session) : Hajime Yoshida, Satoshi Tanaka.
Geant4 Licence
Following a web presentation of Richard Mount, a discussion took place about the liability disclaimer. Maria Grazia Pia pointed the need to have the disclaimer available very quickly in order to clarify situation related to medical activities.
It was noted that the release of a disclaimer need not to be synchronized with the release of a license and a copyright notice. Gunter Folger pointed out the fact that the release of a copyright and or a license will involve modifying of all the source code files and then will induce a new minor release. Guy Barrand and Maria Grazia Pia pointed the fact that the license must enable the use of Geant4 in industry.
The schedule, proposed by Richard Mount, is to send drafts to CB and TSB as soon as possible and seek convergence within one month.
External review
John Apostolakis reported that the CB is arranging the external review, to take place during one week in June. The reviewers will likely consist of a group of seven people : three CBs and four members external to the collaboration.
The timing of the workshop before or after the review was discussed. A preference for holding the workshop after the review was noted.
General Geant4 paper
The schedule proposed by John Apostolakis is to have four steps ; a first review for end of January, a second in March and a third in April. After this the draft paper will be submitted to the review committee. The final step will lead to the submission of the paper after the review.
The main editors are John Allison, Makoto Asai and Katsuya Amako. Ten chapters should be provided. Makoto pointed out some problems with the TeX tools.
The schedule was agreed..
Activity report
John Apostolakis presented a skeleton of the activity report and pointed out that it should be ready for the first of May. The low energy Electromagnetic Group asked for more pages in order to give a fuller presentation.
Collecting statistics on download of Geant4
Some discussions took place around the proposal to request users to give information about their identity and application area in order to download Geant4.
The question whether the system should be in place quickly or a full proposal be discussed was raised. The chair requested a full proposal to be brought to the TSB for discussion.
The Webmaster pointed out that it is possible to obtain simple statistics on user downloads from the Web server.
Experiment representatives noted that we care is required in the interpretation of such statistic, given that users in experiments of centralised Geant4 installations .
Report on Milestones of 2000
Babar need to do some bug fixing and do some comparison with production. Gabriele Cosmo pointed out that some bugs are already corrected.
Katusya Amako reported for Atlas. Mainly debugging is needed. One goal is to have a validation for end of the year. Consultation on choosing hadronic processes would be welcomed. Work on muon distribution had been done ; some differences between G4/G3 had been noticed ; Geant4 give better energy resolution. On hadronics some Y/Z asymetries were observed with Geant4 !
Michel Maire reported on a small set of problems that were identified and resolved in the standard elctromagnetic physics.
Gabriele Cosmo presented software process milestones : a draft of paper has been circulated ; he requested working group coordinators to continue and finish gathering design documents, check diagrams. He also pointed out the need to put in place an architecture sub-group and provide training in tools for project management (Microsoft Project ?).
Training kit milestone
Members welcomed and gave congratulations to the organisers, editors and contributors of the training kits for what has been achieved.
Experiment representatives reported that some introduction chapters were rather promotional; that figures are often not readable. One suggested the idea to have lecture series given at CERN. Also they would like to see more documentation on visualization ; in particular on guidance concerning which drivers are well adapted for what tasks.
It was pointed out that the effort to keep the different parts of the existing documentation up to date is significant. This activity is already time consuming in itself.
Concerning the Geant4 hypernews system: its deployment is still delayed due to man power constraints at SLAC.
There were proposals to study the usage of video for training course and globaly improve existing material; to have more specialized courses, for example on advanced topics in the geometry, on spatial related matters, ... Another suggestion was to try to integrate speaker notes in the tutorial kits.
Keeping track of working group milestones
Some discussions took place about the pertinence of working group milestones and the tracing of the responce to new user requirements. Florence Ranjard proposed to have an action list and report on them at each meeting. John Apostolakis suggested keeping an up-to-date list of development milestones.
Milestones for 2001: discussion of potential goals
The naming of this item needs to be Objectives for 2001.
The chair proposed that draft objectives that are accepted at this stage should be detailed in approximately half a page. This overview should explain the substance of the objective, and nominate a project leader and the parties that will undertake to complete it, and a set of intermediate milestones for its completion. Finally a clear deliverable/fit-criterion is required for each milestone.
1. A development milestone
It was agreed to create a development objective, collecting one to a few objectives from each group.
ACTION 1> Working group coordinators should identify the key goals they commit to implement in 2001, and send a list to the TSB chair by February 7th, 2001.
ACTION 2> A full list of goals for the year 2001 for each working group, including development, testing and support should be created and maintained. This should identify completion times for intermediate checkpoints. First version: February 28st, 2001.
2. Publications
It was agreed that this should be a proposed objective for 2001. In addition to the general Geant4 paper, this should include two other general papers:
- On the architecture of Geant4. Proposed editor: to be found
Contribution: HPW+2 others
Timescale: start July / end fall
- On the Geant4 software process. Proposed editor: Gabriele Cosmo
Draft of detailed objective required.
Action: Makoto to create half a page on this objective (& find editor of architecture paper)
[ Subsequent note: Papers on specific topics written, for example, by working groups or sub-groups are strongly encouraged. A repository for publications and reports should be established, and authors are requested to use it. A list of publications made must be maintained, and a list of publications in preparation or planned could serve as a useful tool. ]
3. Evolution of Documentation and training kit
These aspects should form the support milestone.
Specific goals suggested that could be part of such an objective:
A. An enhanced training kit including speakers notes
B. A video-presentation (approximately 4hr) of an introductory course or a slide presentation synchronised with an audio track.
C. Enhanced documentation with links between user documentation, training kit.
D. Improved tools to aid writers in creating and maintaining diverse documents: user documentation, training kit, exercises,
Detailed proposed objective to be drafted (J. Apostolakis)
4. Comparisons with Data
Agreement to include this in objective of collaboration for 2001. It should include new projects, to be proposed and approved with report deliverables.
A merged objective incorporating these comparisons with data with development of specific application areas is to be drafted by J. Apostolakis and M.G.Pia for consideration of TSB.
The need to improve presentation of results was agreed. A revamped Gallery was seen as a possible method.
5. Interactivity
Given the absence of key people it was unclear whether a separate milestone for this made sense at this time. Development in these areas is important, and can/should be included in development milestone.
6. Regression Suite
A first regression suite was seen as a very worthwhile objective.
Detailed proposed milestone to be drafted by collaboration of physics group coordinators, TSB chair & deputy in coordination with others, eg responsible for analysis.
Action> Preparation of detailed proposal by chair, deputy, physics WG coordinators, +?.
Due: I propose the 28th February.
Issues: Configuration management (dependencies outside & inside G4), test definition, baselining of test inputs and results, change management on reference inputs and outputs, tools for comparisons of outputs, qualification of severity, feedback and archiving of output.
7. Promotion of Geant4 in relevant experimental domains.
Difficult to keep as separate objective. Proposed to incorporate into one objective with Comparison projects (see above).
Objectives:discussion
About regression suite, Maria Grazia Pia proposed to do it in synchronisation with some advanced examples.
Michel Maire pointed out that some coordination between physics coordinators and analysis is needed.
GUI and visualization : seemed to have lost telephone connection with our Japenese coordinators collegues at that time of discussion.
About analysis, Guy Barrand would like to have the interfaces for n-tuples and cloud of points defined by the AIDA group (a workshop at Orsay is planned for that). The objective is to have the implementations working with Geant4 for the end of the year.
Promotion of Geant4 in relevant experimental domains : Maria Grazia Pia would like to see that point as a goal of the whole collaboration. Michel Maire noticed that good part of promotion could be covered by the electromagnetic group. After commenting out that a balance is needed between high energy and low energy electromagnetic group, the chairman suggested that some proposal bee formulated and submitted for vote at a further TSB meeting.
About communication with users, the chairman said that the need is identified and that good part of work should be redispatched to the some support attached to each categories. Some documentation framework need to be developped in order to avoid duplication of writing things. Maria Grazia Pia suggested to approach some professional in the documentation domain. The chairman conclude with the fact that we have to improve training kit and have existing documentation more navigable.
Working group milestones
The chairman asked to coordinators to submit a draft of their main development goals within one week.
Gabriele Cosmo presented the Geometry and Transportation Working Group milestones.
Michel Maire expressed concern about the progress in testing and developing the CAD STEP interface. The geometry group noted this concern and reported on its plans.
John Allison asked about the handling of a problem report from Fermilab concerning propagation in an electric field. The geometry group reported on its activity in responce to this problem report; it noted that this is the top priority of field sub-group.
Michel Maire reported the comparisons and development under way in standard EM working group. He felt the need of the concept of "Cut by logical volume". Some new expertise for high energy processes greater than 100TeV is needed. He would like to see the availability of plotting the cross section (the equivalent of the Geant3 drmat command) ; Guy Barrand said that this could be done in the framework of the new analysis category.
Report on actions from previous meeting
List of actions :
- 1 : Establish an Action list. Belongs to TSB chairperson: J. Apostolakis.
- 2 : Update of the Release Procedure to include the regular review of the Category Diagram(s): G. Cosmo.
- 3 : The review of the User Requirements Document and the creation of a plan for updating it, potentially creating a structured set of URDs for top level and detailed requirements: G. Cosmo.
- 4 : The creation of a repository for storing the results and the raw data of the Assesment. Belongs to Software Management, taken by G. Cosmo.
- 5 : Plan for minor release or patch to update Geant4 3.0 to support shared libraries on Linux and Solaris and updating of system testing to test this configuration. Belongs to Software Management/System Testing: G. Folger, S. Sadilov.
- 6 : Plan for support of the HepMC event generator interface. Belongs to Events Cat., M. Asai.
- 7 : Response to request to specify the lifetime of each instance of B meson. Belongs to Event category: M. Asai.
- 8 : Plan for reduction of initialisation time. Belongs to H. Kurasige, M. Maire.
- 9 : Institutions (experiment, institute, laboratory) representatives are asked to specify what platforms are currently in substantial use for simulation amongst their members.
- 10 : Results of Solaris 7 / CC++ 5.1 certification for Geant4. Belongs to Software Management.
- 1. OK.
- 2. OK.
- 3. Keep.
- 4. Keep.
- 5. Keep.
- 6. HepMC: Keep.
- 7. Hisaya: events & particles: added functionality (it is in the head) for Decay and new dynamic particle.
- 8. UserPhysicsList Saving: Save/store information for materials only today. Physics table cannot be stored yet. Preparing to add it. The possibility to use lazy initialisation of physics tables was suggested. .
- 9. Keep. Send to Gunter
- 10. OK
Youhei Morita commented about HepMC. Concerning B decay (needed by LHCb) something is needed both in particle and in HepMC. The first point is adressed. The second remains to be done.
Regular items
New collaborators :
Two from Maria Grazia Pia (details must be filled HERE). Agreed.
John Apostolakis brought forward the candidature of Gene Cooperman (NorthEastern University), to be admitted to the Common group, to work in the run & events working group on parallelisation. The coordinator seconded. This was agreed pending the conclusion of a formal notification period of two weeks.
New requirement : the CERN/IT representative requested that plans be developed to utilise distributed computing (the grid) in connection with Geant4.
Software management
Gunter Folger would like to move to full direct use of STL (and delete the Rogue Wave wrappers).
He also reported that detailed ISO-compliance warnings will exist in the default compilation on new platforms (g++ 2.95.2 on Linux).
Some problems noticed with installation of the Lizard analysis system ; however the extended analysis example is now included in a CERN binary installation.
Gunter Folger noted that HP and DEC are now being phased out at CERN: the Plus services are now restricted in use and will be switched off at the end of 2002. What happens to Geant4 testing on these platforms (since currently all system testing, including on these platforms, is done at CERN) ?
Gunter requested all institutes to specify the types of platforms that are used for Geant4 by March 15th.
Any other business
Florence Ranjard would like to shorten the TSB meeting. John Apostolakis suggested to try a new formula where the TSB will be fit in one morning session starting at 7h and finishing at 12 h. He pointed out that it will ease situation of remote people.
Next workshop
Will be held in Genova. The possible dates and the constraints were discussed ; the proposals was made to hold it in the week starting on the 1st of July. The chair suggested the creation of a program committee for the workshop and Maria Grazia Pia volunteered to be a member.