Re: Meeting the Scope and Purpose of P754

[Nick Maclaren <nmm1@xxxxxxxxxxxxx>]

Florent de Dinechin <florent.de.dinechin@xxxxxxxxxxx> wrote:
> Background: we are designing portable, correctly-rounded elementary 
> functions, for which reproducibility is simply a must.

That is a matter of opinion.  I am prepared to justify probabilistic
rounding, at least mathematically.  In practice, I agree that only a
few people would be capable of working with it.

In practice, however, I have supported systems with non-reproducible
results from those, and reasonably clued-up users had few problems
provided that the functions were reasonably good.  Reproducibility
is an overstated requirement.


Regards,
Nick Maclaren,
University of Cambridge Computing Service,
New Museums Site, Pembroke Street, Cambridge CB2 3QH, England.
Email:  nmm1@xxxxxxxxx
Tel.:  +44 1223 334761    Fax:  +44 1223 334679