Motion P1788/M0030.01:Level_1_constructors NO
Dear all,
I vote No.
I fully agree with Vincent's argument.
Regards
Jean-Michel Muller
Le 27 févr. 2012 à 12:01, Vincent Lefevre a écrit :
> I vote NO.
>
> My vote would be changed to YES if the following were changed:
>
> The operation num2interval(x) is removed from the motion.
>
> The reason is that it's too dangerous and there are safer alternatives:
> if x hasn't been computed exactly, it will return a point interval,
> thus won't do what the user would usually expect. The most cases where
> num2interval(x) is safe is when x is a simple constant, like a small
> integer, and it would be as simple to write nums2interval(x,x). And
> still if x is a constant, then in general, text2interval(t) is safer
> than num2interval(x): for instance if a constant 1 is changed to 1.1,
> text2interval("1") would simply be replaced by text2interval("1.1"),
> whereas replacing x = 1 by 1.1 in num2interval(x) would be wrong.
>
> BTW, writing nums2interval(x,x) makes more visible than num2interval(x)
> that a point interval is returned, not an interval enclosing the exact
> value represented by x like with text2interval.
>
> --
> 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)
--
Jean-Michel Muller, directeur de recherches CNRS
Lab. LIP, ENS Lyon, 46 allée d'Italie, 69364 Lyon Cedex 07, France
Phone (+33) 4 72728229 - Fax (+33) 4 72728806
Jean-Michel.Muller@xxxxxxxxxxx http://perso.ens-lyon.fr/jean-michel.muller