[Cfp-interest 2432] Re: HAS_SUBNORM paper

[Vincent Lefevre]

On 2022-05-15 06:59:18 -0600, Fred J. Tydeman wrote:
> On Fri, 13 May 2022 15:12:01 -0700 Jim Thomas wrote:
> >
> >What is the purpose of the added parenthetical text in footnote 27?
> >I find the "and/or" confusing.
> 
> To make explicit that flushing applies to both operands and results.
> "OR" in English is confusing: some people read it as exclusive,
> rather than inclusive.

Interpreting it as exclusive wouldn't make much sense here. I'd say
that this is not ambiguous. Also note that the C standard never uses
"and/or".

> >The change to footnote 28 would be substantive if it weren't in a
> >footnote. An implementation following the current footnote might
> >have defined the macros as "absent" where the suggest change would
> >indicate that was wrong.
> 
> With the existing wording, systems the flush results, but not
> operands, are considered absent.
> I contend that they should be indeterminable.

I agree.

-- 
Vincent Lefèvre <vincent at vinc17.net> - Web: <https://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)