Re: [boost] [rfc] Quadrature library for numerical integration ofdefinite integrals

From: Hugo Duncan (hugoduncan_at_[hidden])
Date: 2007-12-06 15:48:25

> Stepping in bravely (where angels fear to tread? ;-) )

But not rushing in, I hope...

> At a very quick slightly-informed glance, this looks a very useful  
> addition to the Boost numeric repertoire,

Sounds like encouragement to me - I shall persevere.

> espcially as it makes
> modern use of C++, including lambda - FORTRAN users won't like it ;-)

Having been forced to read FORTRAN, I have some empathy with them :-)

One of the questions I pondered was whether I should use Boost.Parameter  
or not.  In the end I opted to use it, and I think it works well.  Using  
thirteen parameters does however slow the compiler quite significantly.   
There is also a slight clash with the Math toolkit library in the vault,  
in that it unconditionaly defines BOOST_PARAMETER_MAX_ARITY to a smaller  
value.

BTW, just noticed that math toolkit is in SVN - will it be in the next  
boost release?