NEWS

Version 1.0 :
~~~~~~~~~~~~~

  first public version

Version 1.1 : 
~~~~~~~~~~~~~

  Bugs fixed in kolmogoroff fields:
	- correct destructor
	- define a Bmax() and a _fBmax
  Bug fixed in env1d:
    When a source density is defined then _fXmax is 1.01*Density->Xmax(),
    which gives a non-zero value for a uniform density.
  Bug fixed in particlepropa/photon:
     - Correction of _fTime by adding lTcorr in CheckDetection (particlepropa).
     - Compute EM cascade propagation distance with accurate observer geometry.
  Bug fixed: 
     - slight memory leak in TSophiaInteraction
     - important memory leak in DINT (synchrotron tables)
  Bug fixed in sourcedensity: name of the .fits source density file to 
     be read by cfitsio >= 3.

  Slight improvment to avoid problems in some extreme configurations: 
     - RK integrator modification.
     - secondary pair production photons are propagated only when
     _fTimeStep >= 0.1*IntegratorMinTimeStep
  Slight improvments in configuration files parsing:
    Test number of sources

  Feature added, continuous, uniform source density in 1D:
    Can specify Xmin (in order to easily modelize the effect of the 
    position of nearest source)
  Feature added in the output of photons:
    prints out "Source_Energy" in EeV, in order to know the injection energy
    of the primary nucleon that generated the photon. Mostly useful in 3D
    (in 1D, when only a "grid" of EM showers are propagated, this field is
    arbitrarily set to 0)
  Feature added: test of the 1D approximation for the EM cascade
    if the flag <CutCascade_MagField /> is set within the <Interactions> 
    marker, the following is applied within DINT: At each step all the e+/- 
    with an energy E_e < E_c are removed from the cascade, where E_c is 
    obtained by comparing t_larmor, t_synch, and t_ics. This allows to 
    estimate the error which is made on the point-source flux due to 
    the 1D approximation of the EM cascade.
  New installation procedure: using configure

Version 1.2 :
~~~~~~~~~~~~~

  Tested on MacOSX Intel and SL4 (GNU compiler)
  Small bug corrected, in previous improvments in nucleon.cc
  Bug fixed to compile with MacOS using configure
  
  Slight improvment: a null radio background is implemented; the
    default radio being "Obs"
  Slight improvment: the parameter "A" to cut the EM cascade by comparing
    the Larmor radius and loss timescales can be set directly from the
    XML file.
  Slight improvment: the number of crowns of boxes around which one looks
    for observers in the case of a neutrino may be set by hand to any value
    in the neutrino.h file.

  Feature added: primary photons can be propagated in 1D
  Feature added: possibility to include a probability to follow the
    secondary photons generated by pair production. By following only a
    fraction of these EM cascades and then reweighting the results this
    allows to gain some CPU in 3D configurations.

Version 1.3 :
~~~~~~~~~~~~~

  Slight improvment: use of Hep PDT particle codes in the output to determine 
     particle types.
  Feature added: ROOT output is now possible.