Re: yes on motion 48, with comments
On 2013-08-28 08:05:53 -0700, Richard Fateman wrote:
> I vote yes on motion 48 but with the following comments.
>
> 1. I would hope that any language in which interval arithmetic is
> supported would provide sufficient tools to read and write scalar
> numeric quantities to allow read/write or write/read isomorphism,
I think that this belongs to language standards, and possibly to
IEEE 754 for IEEE 754 number formats (but this is already done
in IEEE 754-2008). P1788 should remain focused on operations
with intervals as arguments or result.
> and that therefore Exact to Interval etc. are unnecessary.
Note that Exact to Interval etc. may still be necessary for implicit
interval types. This depends on the implementation, of course.
> 2. Also .. any language .. would allow reading of numeric quantities
> with rounding up or down, making the issue of enclosure moot.
The enclosure would be ensured only if the numeric quatity is exact
(i.e. the rounding is just due to the conversion from text to value).
This may be dangerous in practice. I prefer the "[x]" notation.
--
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)