I have a question about "Template Name Binding" in Section C.13.8.1 in
TC++PL written by Stroustrup.The example provided is as follow:

void g(double);
template class X : public T {
public:
     void f() {g(2);}   //call g(double);
};
void g(int);
class Z {};
void h(X<Z> x) { x.f(); }   //call g(double). Really??
class Y {public: void g(int);};
//void h(X<Y> x) { x.f(); }   //call g(double). Really??

I tried a test program in VC7 which said has the 100% conformance with
the C++Standard and get a totally different result.

#include<iostream>
void g(double) { std::cout<<"g(double) version"<<std::endl; }
template<class T>
class X : public T {
public:
     void f() { g(2); }
};
void g(int) { std::cout<<"g(int) version"<<std::endl; }
class Z { };
class Y {
public:
     void g(int) { std::cout<<"Y::g(int) version"<<std::endl; }
};
int main()
{
   X<Z> z;
   X<Y> y;
   z.f();  //VC7 result: g(int) version      Not call g(double)
   y.f();  //VC7 result: Y::g(int) version   Not call g(double)either
     return 0;
}

Anyone can give me an advice?

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

MichaelLi wrote:
> I tried a test program in VC7 which said has the 100% conformance with
> the C++Standard and get a totally different result.

According to the docs, conformance of VC7 was "significantly improved"
and that's undoubtedly true and a good thing. AFAICT, they are not
claiming 100% conformance. In fact VC7 does not yet support two-phase
binding of dependent names, as clearly stated in the docs:

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/vclang/html/vclrf1462dependentnames.asp

(sorry for the long link... complain with Microsoft).

Alberto

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]

AlbertoBarbati@libero.it (Alberto Barbati) wrote in message news:<kWhac.101735$FJ6.3741352@twister1.libero.it>...
> MichaelLi wrote:
> > I tried a test program in VC7 which said has the 100% conformance with
> > the C++Standard and get a totally different result.
>
> According to the docs, conformance of VC7 was "significantly improved"
> and that's undoubtedly true and a good thing. AFAICT, they are not
> claiming 100% conformance. In fact VC7 does not yet support two-phase
> binding of dependent names, as clearly stated in the docs: [...]

For what it's worth, they're in good company: of all the compilers
available to me (g++ 3.2, bcc32 5.6.4, como 4.3.3) only como gets it
right.

Cheers,
Nicola Musatti

P.S. Yes, I'm aware that both g++ 3.2 and bcc32 5.6.4 are not very up
to date.

---
[ comp.std.c++ is moderated.  To submit articles, try just posting with ]
[ your news-reader.  If that fails, use mailto:std-c++@ncar.ucar.edu    ]
[              --- Please see the FAQ before posting. ---               ]
[ FAQ: http://www.jamesd.demon.co.uk/csc/faq.html                       ]