The system library seems to define a single value for Inf and Nan of type float (== float64) believe. Will this be cast correctly when used with single precision (float32)? If not, is there a "Nimrod" way to define a generic value?
Thanks.