Re: [OctDev] complex error function

[Jordi Gutiérrez Hermoso]

On 22 November 2012 15:23, Steven G. Johnson <address@hidden> wrote:
> Jordi Gutiérrez Hermoso wrote:
>>
>> With GSL, my suggestion was to actually maintain your code there if
>> GNU libc won't work, but I suspect you might object to the GPL?
>
> I don't have any philosophical objection to the GPL.  However, for
> relatively small packages like this, I find a simple permissive license to
> be more convenient.

Yeah, but if it becomes part of a larger package, I was hoping you
wouldn't dislike it. Good to hear it's not a problem for you. It's a
bit of a pity that Scipy uses the inferior FFTPACK due to GPL phobias.

> GSL also doesn't use C99 complex arithmetic, which is a bit of an annoyance
> in maintaining complex-valued functions in C.   (It seems like none of their
> special functions, aside from trig and exp/log, support complex arguments,
> so it doesn't seem like complex support is a high priority for them.)

I don't think Brian Gough has an objection to C99 or to complex
arithmetic. I just think nobody has done it for the GSL yet. A long
time ago I toyed with the idea of adding complex arithmetic functions
to the GSL and Brian didn't object.

I don't think anyone else has stepped up to maintain GSL with Brian
since the original inception, sadly. If you could work with him,
there's much to benefit everyone.

This could also be an opportunity for you to implement more analytic
or meromorphic functions, if this is your idea of a good time.

- Jordi G. H.