(j3.2006) (SC22WG5.6012) RE: [ukfortran] IEEE_INT

[John Reid]

Malcolm,

Thanks for the explanation. What I had missed was that "round to 
nearest" has been superseded by "round ties to even".

Cheers,

John.

Malcolm Cohen wrote:
> John Reid asks:
>> Do we have a problem with IEEE_INT (see 17.11.11 in N2137)?
>
> The answer to that is clearly NO.
>
> Furthermore, it would be entirely inappropriate to put special wording into a single function out of many.  And even more inappropriate to dump the exact same wording all over the standard.
>
> As it happens, if you want to know what the rounding modes mean, looking in 17.4 "The rounding modes" gives the exact correspondence between our rounding modes (and named constants) and the 60559 rounding-direction attributes, in paragraph 2.
>
> Cheers,
>