BG> Is this valid C++ code?
 BG> 
 BG> int qwe;
 BG> int foo();
 BG> class bar {
 BG> public:
 BG>  int qwe;
 BG>  int ::foo() {return qwe;}
 BG> };
According to my reading of the ISO C++ Standard, the answer is, surprisingly, 
yes.  (I was surprised, at any rate.)  As far as I can see, the reference to 
`qwe' in the function definition for `::foo' should refer to the name `qwe' 
in global namespace scope, not to the name `qwe' in the scope of class `bar'. 
 I deduce this from the fact that none of the name lookup rules relating to 
class scope in section 3 of the ISO C++ Standard apply to the names used in 
definition of a function that *isn't* a member function.
The surprising thing is that a qualified declarator ID in the global 
namespace scope seems to be allowed within a class definition.  I cannot find 
anything in the C++ Standard that disallows this (although I reserve the 
right to find something after further investigation (-:).  Although there's 
nothing logically *wrong* with such a function definition, it does seem to be 
a rather bizarre quirk of the language.
Certainly it is one that many implementors have overlooked.  Several 
compilers that I've tried will fail to compile this code.
 ¯ JdeBP ®
--- FleetStreet 1.19 NR
---------------