Re: Is it g++ bug with "typename"?

[Paul Pluzhnikov]

"Grizlyk" <grizlyk1@yandex.ru> writes:

> I am thinking: why if i will comment lines marked with '#' in the following
> example, all will be compiled, else will not:

You'll get better answers if you reduce your test case (or state
that you couldn't).

> 11.cpp:57: error: no type named 'Second' in 'class
> 11.cpp:57: error: Pxx::Puser::First<int, unsigned int>'

The error message appears to be mis-leading.
Here is what EDG parser reports:

"junk.cc", line 57: error: incomplete type is not allowed
      typedef typename c_First::Second c_Second;
                       ^
          detected during:
            instantiation of class "Pxx::Second<Tobj, c_First, c_Forth,
                      c_Third, Tuint> [with Tobj=int,
                      c_First=Pxx::Puser::First<int, uint>,
                      c_Forth=Pxx::Puser::Forth<int, uint>,
                      c_Third=Pxx::Puser::Third<int, uint>, Tuint=uint]" at
                      line 106
            instantiation of class "Pxx::Puser::Second<Tobj, Tuint> [with
                      Tobj=int, Tuint=uint]" at line 14
            instantiation of class "Pxx::First<Tobj, c_Second, Tuint> [with
                      Tobj=int, c_Second=Pxx::Puser::Second<int, uint>,
                      Tuint=uint]" at line 81
            instantiation of class "Pxx::Puser::First<Tobj, Tuint> [with
                      Tobj=int, Tuint=uint]" at line 141

Cheers,
-- 
In order to understand recursion you must first understand recursion.
Remove /-nsp/ for email.