Re: P1788 1.8 "Language considerations"
On 2012-12-06 16:03:54 +0000, N.M. Maclaren wrote:
> Here is a possible replacement:
>
> Specifically, a language shall specify a set of bindings to the
I would say "a language or implementation" because the implementation
could be just a library.
> operations specified in this standard, with the functionality and
> semantics as specified here. These may be specified as bindings
> to built-in operations, library procedures or otherwise.
>
> Code which is syntactically the same as code operating on
> approximate real numbers, but operating on intervals, shall obey
I don't think this makes sense (perhaps except for OO languages).
And "approximate real numbers" is ambiguous (is an integer such an
approximate real number?). And what about interval operations that
could not correspond to operations on approximate real numbers,
such as intersection, union, reverse operations?
> the following semantics: if the data before and after the unit of
> transformation are regarded as two sets of mathematical intervals,
> the transformed form of all combinations of the elements (the real
> values) represented by the prior set shall be a member of the
> posterior set. A language may specify a mode and constraints on
> programs in which all implementations are required to deliver the
> same results, but is not required to.
I think this should be more general to include constructors,
operations that return a number (such as mid), and so on, i.e. all
the data (whether interval, number, text...) that can appear as
input or output of a unit of transformation.
--
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)