Consider the following code:
#include <iostream>
template <class T>
T v;
template <>
double v<char> = 123.456; // ???
int main() {
std::cout << v<int> << " " << v<char> << "\n";
}
Why I'm allowed (gcc, clang) to specialize template variable for char type as double? What's the correct syntax for variable template specialization?