Have anybody seen Scot Meyers' video on Universal references? In order for a reference to be "universal" two requirements should be fulfilled :
1)variable should have double ampersand after it's type : sometype&&
2) the type should be deduced .
The second point is somewhat confusing to me: despite examples provided there, Meyers hasn't given a clear definition :
what does mean deduced type?
So the question is : could you give if not a clear difinition of what criteria a type should fulfill in order to be called "deduced", but at least your vision on it.