Contributing to GMP

[Anders Andersson]

On Tue, Feb 18, 2020 at 1:09 PM Michele Guerini Rocco
<rnhmjoj at inventati.org> wrote:
>
> Hello everyone,
>
> I have written, as an exercise, a program to compute the Euler-Mascheroni
> constant (γ) to arbitrary precision using GMP. It's not meant to beat
> existing programs (like y-cruncher) or anything like it, but it works.
> Anyway, I had to compute the natural logarithm of a natural number to
> arbitrary precision, so I wrote a `mpq_log_ui` function to do that,
> something which GMP doesn't provide.
>
> Since this function could be useful in general, for more pratical
> purposes that computing a constant, I thought about contributing it
> back. I'm a layman in numerical analysis and I'm not sure the
> implementation it's up to the GMP standards but I can work on it.
> So how can I contribute to GMP?

Does it really make sense for this to be a rational number function
instead of an mpf_log_ui? Everything else dealing with rational
numbers works with *exact* numbers, and it seems to me that a log_ui
will rarely be exact.