Paul A Bristow wrote:

...

Following the successful review of the math-toolkit of special functions and statistical distributions the main feature request was for a better way of customising the library: both for choosing between precision vs speed tradeoffs, and for determining how errors are best handled. I've been experimenting with various policy-based interfaced based on Boost.Parameter, and I think I now have something useable, so I'd like to know what people think:

This all looks *very* (even excessively!) cool, and has real practical value.

But it is more complicated - though most people just use the default defaults ;-)

I see the error messages as being *really* important, so some other mechanism than BOOST_CURRENT_FUNCTION must be devised to avoid the obfuscating pages added by MPL. More MPL ;-)

Yep, we could just specify the function name manually "boost::math::tgamma" etc, I guess we could maybe even use Boost.Format to insert the template type - I think all the error handlers get passed that type separately anyway - I'll have to check that.

Finally, I wonder what the effect on the C++ ISO Standards process will be? Will WG21 take fright? Do we care?

Not at present :-)

PS It also has a cost of some hacking hours ;-)

I know :-( John.