Re: [boost] type_traits yes/no_type

From: John Maddock (jm_at_[hidden])
Date: 2002-11-08 07:22:53

> I notice that our type_traits headers still define:
>
>    typedef char yes_type;
>    typedef double no_type;
>
> Which is non-portable w.r.t. sizeof() differences. Is there any reason
> we can't change no_type to
>
>    char (&)[2]
>
> or
>
>     struct { char[2] x; } // if it needs to be constructed

I seem to remember that floating point types had a guaranteed minimum size,
but I can't see it in the standard, so I'll change that as suggested.

John Maddock
http://ourworld.compuserve.com/homepages/john_maddock/index.htm