Re: Question about interpretation of ISO C++ standard

From: rwgk_at_[hidden]
Date: 2001-05-22 14:19:29

Thanks for the hint!
As a test, I removed the semantics parameter. The online gcc
compiler then complains about foo1 not being declared.
This is my first piece of evidence that Koenig lookup
actually exists!
Could one of the C++ language authorities please confirm
that foo1 ought to be found in the namespace of the
semantics parameter? I'd like to be a bit more certain
before sending a bug report to Compaq.
Thanks!
        Ralf

--- In boost_at_y..., Larry Evans <jcampbell3_at_p...> wrote:
> rwgk_at_m... wrote:
> 
> > Is the following piece of code supposed to work with a fully
> > conforming
> > compiler?
> 
> ...skipping
> 
> >
> > Here is what I find:
> >
> > gcc-2.95.2: works, but only because foo1 is instantiated in the
> > global namespace. This behavior is usually considered a bug.
> 
> It compiles without error on g++'s online compiler at 
http://www.codesourcery.com/gcc-compile.shtml