Re: That other flavour...
On 2013-09-20 17:15:52 -0700, Dan Zuras Intervals wrote:
> I doubt it was Michel's intention to raise this issue but I will:
> Can we (or should we) restrict implementations to 754-2008?
There's no reason for such a restriction.
> Given its wide acceptance throughout the world, it would not
> restrict our customers much.
It could for various users, e.g. in the context of multiple precision.
Or if one wants to use rational arithmetic or some other arithmetic
(possibly based on IEEE 754, like double-word...).
Moreover, even if the processor supports IEEE 754, the compiler
may not necessarily to support it entirely. For instance, Clang
currently doesn't support it at all. Or the user may want to
disable some IEEE 754 features (e.g. because they prevent some
optimizations that could be proved as safe in the user's program)
while keeping interval arithmetic.
--
Vincent Lefèvre <vincent@xxxxxxxxxx> - Web: <http://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)