- missing initialisation of m_ReactionRegion in Actar.cc

nps
        - the flag is --random-seed

        - The state of the G4 random generator was never cleared
        - bigger and bigger data were written to the tree

        - at the beginning of each event, in the TTree
        - at the end of the simulation in a TAscii file

        - The elastic process could trigger to daughters particle in
        certain cases
        - This lead to an infinite chain of particle generation stopping
        the programm
        - Beam reaction now test the parent particle have trackID 0

        - this is within an effort to allow out of source build

Fixing conflict