Consider switching to use _t types.

[externl]

I think we should consider switch our numeric types to use _t types (Is there a better name for these?).

• https://en.cppreference.com/w/cpp/types/integer
• https://en.cppreference.com/w/c/numeric/math/float_t

Old Type	New Type
typedef unsigned char Byte	uint8_t
typedef short Short;	int16_t
typedef int Int;	int32_t
typedef long long int Long;	int64_t
typedef float Float;	float_t
typedef double Double;	double_t

[bernardnormier]

We should check how compatible this proposed change here with the CI platforms.

[externl]

Did a preliminary test on macOS. Updated the compiler to generate these types and changed nothing else. Going to test MSVC and GCC next.

[externl]

So, macOS and Windows work, however Ubuntu's GCC typedefs int64_t to unsigned long, whereas we use unsigned long long.

So while this is not a backward compatible change I think it's still a change worth making. Yes, users that are upgrading will have to make some changes. I think that's fine as it will bring us more inline with what modern C++ code should look like, as I think these type are the expected norm now.

https://google.github.io/styleguide/cppguide.html#Integer_Types

[pepone]

I agree on updating this to use _t types.

however Ubuntu's GCC typedefs int64_t to unsigned long

I assume they would use long int in 64bits platforms and long long int in 32bit platforms see https://stackoverflow.com/a/60391237/491807

[externl]

This was done in #1806

[externl]

In the end we only did with fixed with types, so we did not use float_t or double_t.