Re: math constants

[Ed Smith-Rowland <3dw4rd at verizon dot net>]

On 02/04/2013 01:47 PM, Paolo Carlini wrote:
> Hi,
>
> On 02/04/2013 06:25 PM, Ulrich Drepper wrote:
> > On Mon, Feb 4, 2013 at 12:16 PM, Paolo Carlini 
> > <paolo.carlini@oracle.com> wrote:
> > > The general idea makes *a lot* of sense of course. Before adding too 
> > > much
> > > code, however, I think we should double check whether something is 
> > > going on
> > > already in terms of a facility planned for TR2 or even existing 
> > > already in
> > > eg, Boost.
> > boost has something incredibly ugly.  They use template functions to
> > return the values and some ugly way to allow initialization of data
> > types other than float/double/long double using strings.
> Ok...
> > I think constexpr are much more in line with C++11 and going forward
> > and the math_constants class follows the numeric_limits approach.  If
> > you could look through the meeting notes that'd be appreciated. I
> > wouldn't even know where to start.
> Agreed. I'll check and also ask people at the next Meeting. In the 
> meanwhile we can certainly add something which we think makes sense in 
> terms of interfaces, uglified for now, to be safe.
>
> Paolo.

I just remembered I found it necessary to add math constants for the old 
TR1 math functions.

If you look at include/tr1/special_function_util.h I had a 
floating_point_constant template in analogy to the standard 
integral_constant.
Then I added a bunch of math constants.  So we have some prior art in this.

I would be in strong support of a fully standard solution along these lines.

Ed