Do not set particle lifetime to 0 (#5472)
An attempt at solving #5472 (closed)
If a particle has a lifetime of 0, the calculation here will result in NaN. On some platforms (at least MacOS and FreeBSD) it leads to a crash here.
This MR set a minimum value to std::numeric_limits<float>::epsilon()
instead of 0.0f
.
I don't know anything about NIF files, so maybe a value of 0 has a special meaning that should be handled differently?