Minutes of MCNSI meeting at PSI, 2/10-2006

2. Draft, KL 6/12-2006

Present:
Uwe Filges, Heinz Heer, Geza Zsigmond, Jan Saroun, Mark Hagen, Peter Willendrup, 
Kim Lefmann, Phil Bentley, Klaus Lieutenant, Emmanuel Farhi, Sasha Ioffe

Next meeting:
Prague, spring 2007 (JS)

0. We should try to advertise MCNSI at ECNS-07. E.g. by holding a MC tutorial.

1. Status from the groups
  - NPI: Has hired Jiri Svoboda for 7 months to work on GUI (period already expired).
         Some money is still left
  - PSI: No change in personnel. Has used 2/3 of the recently allocated funds.
  - Risų: Peter Christiansen has left the McStas group. New person will be hired soon.               Polarization has been implemented in McStas. Release 1.10 due in late 2006.
  - HMI: VITESS is now under open source, GPL licence. Version 2.7 is ready for release.
         Carlos Fehr works until Nov. 2006 with collecting examples of past
         simulations. Manual is in preparation.
  - Kjeller: KLi will spend most time on hardware, will simulate powder diffractometers.
  - ILL: A permanent position is open in the software group. Many trainees are working
         on McStas.
  - FZJ: Sergey Manoshin is working effectively on a) polarization with time-dependent
         fields. b) Neutron focusing lenses.  
  - (USA): The DANSE project has been funded with 10 M$. Will work on composite samples.  

2. (I HAVE NO NOTES ON THIS ISSUE)

3. Discussion on software developments in the FP7 version of NMI3 before the
   strategy meeting in Taormina, 9-13 October. 
   (This point took more than two hours, and ended with a common understanding, but
    without a single "decision text". The most important views are collected below.)
  - AI: We must maintain the packages, they are needed to develop new instruments.
        Dubna should be partner in FP7 simulations
  - EF: FP7 will last 7 years, we cannot spend all time just developing packages.
  - AI: We should work on more realistic descriptions of instruments.
  - EF: If we should join data analysis we need a strong development. We must approach
        realistic virtual experiments.
  - UF: It will not be good if the money for simulation work are split between different
        JRAs
  - EF: We should keep the simulators together, because we speak the same language.
  - PW: We can still collaborate even if we are in different JRAs.
  - AI: We could overlap with other JRAs by common milestones, common employees.
  - MH: You would need a simulation forum to maintain synergy.
  - PW: We could have one simulation JRA but with stronger connections to other JRAs
  - AI+KL: MCNSI should not be a closed club, we must promise to give input to others.
  - EF+KLi: We should avoid a situation where we are just subcontractors to other JRAs
  - HH: We can promise to do detailed virtual experiments and support data analysis
        activities.
  - EF: The work that is needed will be performed. We cannot write down everything from
        the start. It will be difficult to be member of more than one JRA.
  - PB: We need something new, a rebranding. "advanced virtual neutronics", instruments
        for ESS long-pulse target, test of data analysis packages.
  - AI: We should specify tasks that can be fulfilled within reasonable time. After that
        we can work freely.
  - EF: We can promise: Better accuracy, realistic samples, link with data analysis,
                        polarization, UCN, correlate data with simulations
  - HH: We must realize that other JRAs are not likely to give ressources away.
  - MH: If the simulation ressources are distributed, then competences will be diluted
        and may disappear. You should also know that a number of things are being
        developed at SNS.
  - UF: We must have a stand-alone simulation project
  - (all): Suggestion for simulation JRA: 
            - Virtual exp. for data analysis
            - New algorithms, e.g. for new optical components
            - Instruments at new sources
            - Teaching (schools, university)
  - AI: The trends and needs are:
            - Focusing guides
            - Polarization
            - Time dependence
            - Global optimization
        The simulations should not be swallowed by data analysis. We need to have support
        from the facilities for simulation activities.