Re: Motion P1788/M007.01_NaI: Discussion period begins
On 2009-08-06 12:18:16 +0200, J. Wolff v. Gudenberg wrote:
> Vincent Lefevre schrieb:
> >2.1 says: "According to motion 3, if both arguments are finite numbers
> >and l ≤ u, or l equals −∞, or u equals +∞ that constructor returns the
> >F interval [l,u] otherwise NaI is returned."
> >
> >You should also exclude the cases l = u = −∞ and l = u = +∞.
> These are excluded since u=-oo as well as l=+oo is not allowed
They are not allowed in the interval representation, but these values
can occur in the constuctor. One needs to explicitly say what happens
in this case (just like it is said l <= u for finite numbers).
Something like:
According to motion 3, if l < u or if arguments are equal finite
numbers, that constructor returns the F interval [l,u]. Otherwise
NaI is returned.
Also, I think that there should be a level-2 specification and
a level-3 specification (the text seems to mix both). Level-2
specification is clear. But for level-3 specification, the encoding
of intervals has not been specified yet (in particular concerning
the signed zeros).
--
Vincent Lefèvre <vincent@xxxxxxxxxx> - Web: <http://www.vinc17.org/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.org/blog/>
Work: CR INRIA - computer arithmetic / Arenaire project (LIP, ENS-Lyon)