-----------------------------------------------
 Changes in v2.2.2 compared to v2.2.1
-----------------------------------------------
bayliffe (2):
      Force single precision read of density from sphNG dumps (as it is stored) even when compiling splash as double precision; otherwise the values returned are junk.
      Added calc amp to give the amplitude rather than just the peak to peak difference; differs by factor of a half.

dprice (37):
      bug fix with reading real4s from sphNG files if splash compiled in double precision
      ndspmhd read handles one-fluid dust arrays (iformat=5)
      renamed read_data_dansph->read_data_ndspmhd
      bug fix with dustyshock + isothermal eos
      SPLASH_TO_GRID env variable can be used to specify particular columns in splash to grid
      SPLASH_CENTRE_ON_SINK environment variable added; works for ALL data reads
      minor change to help info
      docs added for calc diff and SPLASH_CENTRE_ON_SINK
      added some more docs about SPLASH_CENTRE_ON_SINK
      Implemented particle tracking by type; can be saved to splash.defaults
      compiler warning (type conversion) fixed in calc diff
      plot_pap interface includes optional units argument
      pre-calculation of one-fluid dust quantities added
      bug fix with read of Kdrag from ndspmhd .in file
      added fictional dust particle creation for one-fluid dust visualisation (NSPLASH_TWOFLUID=yes)
      output from splash calc to screen to many more decimal places
      added data read for Aly Reheam (splash-aly/make aly)
      bug fix with reading gadget initial conditions files; also only looks for Nh, Ne if iFlagCool=1 instead of iFlagCool.ne.0
      BUG FIX with overlaid ticks not appearing on first panel pdf/eps/ps plots (thanks to M. Bate)
      added ability to plot second y axis to setpage routines
      bug fix with alt y axis stuff
      bug fix with length of alt y label
      added alternative y-axis as option in axes menu
      preliminary work on interactive switch-to-movie mode (commented out)
      added extra -lX11 flag to link step; seems to solve issues with XFlush+Ubuntu
      bug fix with iaxis=4 (re-scaled y axis) if transformations applied; now works OK
      sphNG read a bit less verbose
      brightness correction NEVER applied for opacity-rendering (until I fix it)
      bug fix preventing compilation with gfortran 4.6
      one fluid stuff uses dust fraction instead of dust-to-gas ratio
      unused variable warning fixed
      unused variable warning fixed
      Toomre Q plots use actual speed of sound if read from dump file (if ispsound=column containing sound speed)
      updated docs for v2.2.2
      v2.2.2