[boost] Re: Type Traits, Take Two

From: Andy Sawyer (boost_at_[hidden])
Date: 2000-01-14 17:37:24

On 08 January 2000 03:17, Gabriel Dos Reis wrote:

> "Andy Sawyer" <boost_at_[hidden]> writes:

> |  Not to mention bool (I still maintain bool is neither signed 
> nor unsigned.
> 
> I fully undertood you point, even when I disagreed ;-)

 I seem to remember that you agreed on this point (or, at least, didn't
_actually_ disagree) - albeit for different reasons :-).

> | I am actually of the opinion that the same is true of char, but for
> | different reasons.)
> 
> As numeric_limits documents the implementation defined aspects of
> fondamental^W arithmetical types, I think that
> numeric_limits<char>::is_signed is implementation defined.

 I believe that you are correct (again:-). However, my reasoning for
considering char a type-without-signedness is more philosophical than
standardese.

> BTW, IRC you raised the issue of what ot means to be signed.

 Was that during the great numeric_limits<bool>::is_signed debate?

Regards,
 Andy
-- 
Andy Sawyer, Technical Director, Sufficiently Advanced Technology Ltd.
mailto:andys_at_[hidden]        ICQ:14417938        http://www.morebhp.com
Mobile: [phone] [Voice/SMS] ([phone] [Fax]
"Any technology distinguishable from magic is insufficiently advanced"