This release allows configuring the physics model for tau energy losses and for neutrino DIS interactions. In addition, it patches some bugs, and it adds Python3 compatibility.
- The physics model can now be configured using the `danton_physics_set` API function, or the corresponding data card option. For DIS, the user can switch between CSMS ([Cooper-Sarkar, Mertsch and Sarkar](https://arxiv.org/abs/1106.3723)), BGR18 ([Bertone, Gold and Rojo (2018)](https://arxiv.org/abs/1808.02034)) or LO (Leading order computation, following [Gandhi et al., (1995)](https://arxiv.org/abs/hep-ph/9512364)). By default, BGR18 is used. Note that in previous versions of DANTON, the LO computation was used . Note also that, with this new version, the kinematic is still simulated using LO expressions, even when using CSMS or BGR18 cross-sections. Only the total cross-section is rescaled. Corresponding PDFs are used, however, depending on the selected DIS physics model.
- The "requested" counter has been patched. Simulations should deliver a proper number of events, as requested by the user.
- The forward flux end condition has been patched. In forward mode, all particles should now properly stop when reaching the sampling surface. In previous versions, in some rare cases, secondary taus could stop before reaching the sampling surface.
- The recorder "verbosity" option has been removed, since it was not implemented, actually.
- The generation weight of forward Monte Carlo events has been patched. Forward and backward cases now deliver consistent weights. Previously, the forward Monte Carlo weight did not include the generation solid angle.
- The `danton.py` utility module now supports both Python 2 and 3.