Re: requiring hardware is futile
On Wed, Nov 19, 2008 at 3:02 PM, Ian McIntosh <ianm@xxxxxxxxxx> wrote:
> Interval Arithmetic is clearly important for some applications, but I'm
> skeptical that mainstream processors will make major changes for it. The
> standard will see greater acceptance if it is reasonably implementable with
> no changes to existing CPUs, languages and compilers, and better
> implementable with just minor changes. I think that's doable.
At the risk of sounding like a broken record, an interval standard
that does not go out its way to require hardware support has higher chance
of finding realistic bindings for programming languages -- at least the ones
that are in wide use or have international standards. The Interval Standard
should allow for the "as-if rule" to apply, so that compilers can
provide adequate
support.
-- Gaby