Does it actually happen?

[L B]

http://ideone.com/T0egOt

#include <iostream>
 
namespace ns //::ns
{
    void f() //::ns::f
    {
        std::cout << "::ns::f"<< std::endl;
    }
    namespace ns //::ns::ns
    {
        void f() //::ns::ns::f
        {
            std::cout << "::ns::ns::f"<< std::endl;
        }
    }
}
 
using namespace ns; //::ns
 
int main() //::main
{
    ::ns::f(); //::ns::f
    ::ns::ns::f(); //::ns::ns::f
    f(); //::ns::f
    ns::f(); //error, ambiguous
}

Obviously this is another reason that "using namespace <something>;" is bad, but someone I showed this to told me it never actually happens in real code that you have something with the same name as the thing it is inside of (besides constructors/destructors).

They even ignored me when I said a namespace may have a class in it that has the same name, which (I thought) happens quite often in programming. Who do you think is right?

[Cubbi]

I don't think it's too much of a a stretch: one of the classes I'm using a lot is named the same way as the (4th-level nested) namespace it's in, except for a minor capitalization difference. And yes, the naming scheme actually makes sense (and no, I wasn't the person who named it). I don't think it's "quite often" though.

[Framework]

LB wrote:

"I showed this to told me it never actually happens in real code that you have something with the same name as the thing it is inside of (besides constructors/destructors)."

...but you've shown him/her an example of it. Real-world code is no different than code that one has written for their personal use. If I wrote an array wrapper for a project at work, it would functional exactly the same if I used it in a personal project.

The fact is, it will only happen if the program/library is poorly structured or if the naming convention is appalling. Programmers who use "using namespace ..." excessively are likely to trigger such ambiguities.

_Wazzak