- NPAnalysisFactory was using the detector factory preprocessor
        variable to avoid multiple inclusion
        - Fixed using proper pre processor variable
        - the bug was preventing the compilation of analysis process
        with some detector, behaviour not really defined (depend on
            order of inclusion

! Reading 1 TH2F instead of 2 TH1F

! angular ditribution from an TH1F histogram to simulate p, d, t...

        - Read sensitive is now call using pointer : 8% faster
        - Generate event now call via pointer: marginal 1% faster

* the distribution

        - Obsolete TOF scorer got now threshold
        - Trapeze issued a warning of Time size different from E size

        * Issue was in NPeam deleting histo before root tries
        * also pass a couple of variable to the heap instead of the stack

        - Negative strip could be obtained leading to seg fault

        * Still return 0 for life time using < or >
        * work with unit correctly otherwise

        Can now deal with nuclei that do not exist

        - Skip the missing include depending on G4 version

* energy loss table name