[cfp-interest 3817] Re: CFP Future

[Paul Zimmermann]

       Hi Damian,

very interesting!

> Some thoughts.
> 
> a) Short term, move all mention of Imaginary arithmetic to another Annex
>     or whatever mechanism seems good - see (e) below
> 
> b) Cater for changes to IEEE 754 2029
> 
> c) Will we need to cater for P3109 a,k,a, '_Float08' into the future
> 
> d) Learn from experience with _FLOAT16 and address whatever issues arise
>     from there.  We already have a problem, that of the INFINITY macro in
>     an implementation which supports _FLOAT16. See CFP 3746.
> 
> e) Long term, once IEEE 754 2029 gets an Addendum on Complex and Imaginary
>     types, address this. Note motion for such an Addendum is not yet put
>     due to conflicting work schedules.
> 
> f) Ongoing tweaks to wording to improve readability of, and tightness of
>     language within, the C standard
> 
> g) Integrate any TS 18661 features in the maths libraries that are
>     deemed required by WG14
> 
> h) Remove erf(), erfc(), lgamma() and tgamma() from the math library as
>     they really are too specialized and belong elsewhere.

on the other hand, would sincos belong to the math library?

Paul