I've seen some code where they don't use primitive types int, float, double etc. directly.
They usually typedef it and use it or use things like
uint_8 etc.
Is it really necessary even these days? Or is C/C++ standardized enough that it is preferable to use int, float etc directly.