pub fn min_float(a: f32, b: f32) -> f32
@return the minimum of two floats Matches the C ternary exactly, including NaN propagation (a < b is false for NaN).
a < b