process PID to avoid collision when running multiple npanalys in
  parallel

        - recording not yet done
        - add appropriate method to EventAction
        - add correct option to NPOptionManager

        - The fIsSplit variable was not initialised to false

        - output one tree per detector in a folder with associated .tree
          file

        - User can build Friends entry by hand
        - User can use .tree file as entry point to automatically
          generate friends input
        - User can use -T flag with .tree file

nps
        - the flag is --random-seed

 - P flag is used to specify the port

* Fixing pile up in silicon scorers

        - Readin file without RunToTreat file
* Suppress using namespace std from Core header

not just the total number.

Also fixed minor bug - in the event loops the loop variable was unsigned int.
Really it should be unsigned long.

        - However does not check that the provided file make sense for
          G4.

        - correct spelling
        - streamlined presentation

* old one
* Run replace.sh to replace old matta@ipno.in2p3.fr address

* updating copyright
* updating a matta email

* option is given

- Adding test to travis CI

        - User need to provide a macro path with B flag
        - No Vis or GUI macro are executed in batch mode

        - Macro support any command except exit
        - exit cause the macro to stop, not Geant4
        - I don't have any fix for this yet ;)

        - gcc get confused by global variable and do not mangle the name
        correctly
        - fixed by removing the global varaible in question to
        NPOptionMnager

* Adding version number display

* use full for fast debugging and fast analysis

  - CMake compilation
  - PlugIn API for detector
  - Dynamic Loading of detector library at RunTime

+ Remove NPTagManager class

 - VDetector now have SPectra related classes
 - Detector Manager take care of calling the associate method
 - Still, the writing of the histogramm classes is to be done

 - The analysis project is in pMakePhysicalTree
 - The macro is launch putting the usual NPTool option in the macro string argument
 - Parrallel computation and merging of the output tree in memory (not suitable for very large run...)
 - Some change in the ROOTOuput to fit the requierement of pROOF, using the NPTOptionManager option --proof to activate them