15 Dec
2004
15 Dec
'04
10:56 a.m.
On Tue, 14 Dec 2004 12:30:55 -0700, Jonathan Turkanis <technews@kangaroologic.com> wrote:
I agree with many of your points, except the main one: that Boost.Rational with built-in intergal types is useless without rounding.
if you are really, really convinced that this is the case, there is a code snippet at http://www.ccg.hu/pub/src/rat.cpp -- it shows that the current finite precision boost::rational semantics can be implemented a lot more efficiently
- round to the nearest representable value I think you are arguing that ... is easy to compute, and should be the default. Right?
yes br, andras