PRNG is a replacement for the default Random::PCG32 stdlib prng that implements
the xoshiro algotithms to generate pseudo-random UInts in 64-bits.
Performance is comparable, but quality of generated Floats is much better.
Inherit this and implement #next_u, #next_f and #jump to build you own generator.
Instance methods
next_f32
Generate a uniform-distributed random Float32 in [0, 1).
next_f64
Generate a uniform-distributed random Float64 in [0, 1).
next_i32
Generate a uniform-distributed random Int32 in range Int32::MIN..Int32::MAX.
next_i64
Generate a uniform-distributed random Int64 in range Int64::MIN..Int64::MAX.
next_u32
Must return an uniform-distributed UInt32.
next_u64
Must return an uniform-distributed UInt64.