Re: [boost] is_union implementation

From: Howard Hinnant (hinnant_at_[hidden])
Date: 2002-10-16 08:58:21

On Wednesday, October 16, 2002, at 05:32  AM, Aleksey Gurtovoy wrote:

> On a related note, it seems to me that the current semantics of
> 'boost::is_class' trait:
>
> <docs quote>
>     "Evaluates to true only if T is of class/struct type."
> </>
>
> does not match the standard - section 9 clearly states that union is a 
> class
> too.

Section 3.9.2/1 just as clearly gives union a classification distinct 
from class. <shrug>

The type traits lib used 3.9 as the guide, and I think a distinction 
between class and union is useful in practice.

-Howard