Re: Testsuite failures of gawk 5.2.0

[Eli Zaretskii]

> From: Wolfgang Laun <wolfgang.laun@gmail.com>
> Date: Fri, 9 Sep 2022 10:03:48 +0200
> 
> On Fri, 9 Sept 2022 at 07:46, Eli Zaretskii <eliz@gnu.org> wrote:
> 
> > >
> > >       double d = -sqrt(-1.0);
> > >
> > > Sheesh, what a headache.
> > >
> >
> 
> It's irrational, not a "headache". 😉

No, it isn't.  A NaN is "not a number", so the rules of using it are
"different".  E.g., it's the only FP value that fails both equality
with itself and inequality with itself.  Some might consider that
"irrational".