stdint: no love for C++?

So stdint and exact-width types are part of the new C standard, but not C++. Fortunately, I can still #include <stdint.h> (although not portably); however what makes least sense is that, although the XintN_t types are made available, their counterpart max sizes, XINTN_MAX defines, are being held back by an "#if !defined ( __cplusplus) || defined (__STDC_LIMIT_MACROS)"

Why is this? Why allow C++ to include the types, but not their max sizes? What would be the danger in removing the ifdef from the header? Or re-defining the macro in my own code?

Because C++ has its own way of handling limits. Its nothing personal, just an attempt to maintain uniformity within C++. Not uniformity across the two languages.

I've heard that the next C++ standard aims to aspire to the very same... but in either case, how do I go about this:

Code:

---

uint64_t max_val = UINT64_MAX

---

in C++?

Since you know by definition this had better be a 64-bit type, 0xffffffffffffffffull?

Or (and also I didn't try this one either) maybe numeric_limits<unsigned long long>.max?

1) Boost contains <boost/cstdint.hpp>. It includes <stdint.h> where available, emulates it where the underlying types are available, and you're screwed elsewhere anyway.

2) Use std::numeric_limits. <stdint.h> only contains typedefs for types that the compiler supports anyway, and thus there's a very good chance that numeric_limits is specialized for them. (By the way, max is a function. Thus, std::numeric_limits<uint64_t>::max(). Also, beware of the min() and max() macros in windows.h.)

Quote:

---

Why is this? Why allow C++ to include the types, but not their max sizes?

---

Interesting question - but one better asked at the support forum of whatever compiler does this.

Thanks, I'm using "std::numeric_limits<uint64_t>::max()" now.