- putting own header file first
  o for reason see http://www.cplusplus.com/forum/articles/10627/#msg49679 and
                   http://www.cplusplus.com/forum/articles/10627/#msg49680

- remove 'using namespace' from header files
- remove 'using std::string' and 'using std::abs' from header files

For reason see:
o http://stackoverflow.com/questions/5849457/using-namespace-in-c-headers

- updating .gitignore

For ParallelT method: If BoundaryGeometry is allocated, use it for the iterative solver in a field free region

Fixed small issue with singleParticleDump not working correctly in CyclotronTracker in RK-4 mode, made combined LF-2/RK-4 tracker default

fix problem when many sections are glued together

- improve check if sections are glued together

*) Fixed a problem with saving the lost particles if not all processors have lost particles. \n*) Consolidated Tracker_RK4 and Tracker_LF in ParallelCyclotronTracker. Enclosed in #ifdef for now. Activate by uncommenting #define GENERICTRACKER in ParallelCyclotronTracker.h. \n*) Added the general OPAL OPTION boundpdestroyfreq for Cyclotron Tracker.

Cosmetics, Typos and a bugfix in calculateBeamParameters_cycl() where the energy would always be 0 and thus the emittance inf

xpang: fixed a bug in CSRWakefunction.cpp and added CSRIGFWakefunction.cpp for the integrated green function method

Injector2 runs are terminated normally now. MPI_File_write_at function was called after MPI_FINALIZE

Delete object of abstract class to finalize the run properly -Use smart reference-counted pointer class for dynamic memory management to avoid memory leak

Fixed some issues with new asignment operators in variable RF cavities that prevented gcc from compiling on merlin PLEASE DOUBLE-CHECK

1. Changed the way eKin_m is calculated in PartBunch to be consistent between P and E especially when restarting, 2. Removed some debug outputs.

Cosmetics, Typo, and an indexing error in Distribution.cpp that lead to (n_processors - 1) extra particles

Major Changes: 1. No more dumping or restarting in LOCAL frame. 2. OPAL-CYCL now recalculates reference* variables before dump and saves in H5 file. 3. OPAL correctly reads reference* values and calculates beta*gamma correctly from Ekin (was factor 1000 wrong before) Note: Transformation of bunch data into local frame must now be done in POSTPROCESSING -> upgrade H5root

Some refactoring and more comments for computeSelfFields_cycl(), fixed missing gamma factor for electric field -DW

No commit message