accuracy of the epsilon when moving to next cell in routine propagate
still unclear, not documented, and some parameters are hardcoded... possibility to make it adaptive?
still unclear, not documented, and some parameters are hardcoded... possibility to make it adaptive?