(Yes, I'm replying to myself as I forgot something. Sorry :)

Daniel Frey wrote:
>  
>  I think it depends on the compiler. Compared to the canonical
>  
>  number operator++( int )
>  { number tmp( *this ); ++( *this ); return tmp; }

This should of course be:

const number operator++( int ) { ... }

to prevent i++++; to be legal. Maybe we should even change the define to
prevent this kind of problem like this:

>  #ifdef BOOST_COMPILER_LIKES_POSTFIX_INCREMENT_OPTIMIZATION
>  
>  template< class T > struct postfix_incrementer : public T
>  {
>     postfix_incrementer( T& rhs ) : T( rhs )
>     {
>        ++rhs;
>     }
>  };
>  
>  template< class T > T postfix_increment( T& rhs )
>  {
>     return postfix_incrementer< T >( rhs );
>  }
>  

#define BOOST_POSTFIX_INCREMENT( T ) const T operator++( int ) { return postfix_increment( *
this ); }

>  
>  #else // BOOST_COMPILER_LIKES_POSTFIX_INCREMENT_OPTIMIZATION
>  

#define BOOST_POSTFIX_INCREMENT( T ) const T operator++( int ) { T tmp( *this ); ++( *this )
; return tmp; }
>  
>  #endif // BOOST_COMPILER_LIKES_POSTFIX_INCREMENT_OPTIMIZATION
>  
>  Now implement operator++( int ) like this:
>  
>  number operator++( int )
>  { BOOST_RETURN_POSTFIX_INCREMENT( number ); }

We needn't "implement" it any more, we just "add" it by a single line to
our class:

BOOST_POSTFIX_INCREMENT( number )

and that's it. I also took this approach for other operators, but this
new optimization technique needs to be merged. Iif I find some time,
I'll check it in in a few days...

Regards, Daniel

--
Daniel Frey

aixigo AG - financial training, research and technology
Schloß-Rahe-Straße 15, 52072 Aachen, Germany
fon: +49 (0)241 936737-42, fax: +49 (0)241 936737-99
eMail: daniel.frey@[...].de, web: http://www.aixigo.de
_______________________________________________
Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost