Re: [OctDev] complex error function

[Steven G. Johnson]

Jordi Gutiérrez Hermoso wrote:
> 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.

GSL does have complex arithmetic functions:
        http://www.gnu.org/software/gsl/manual/html_node/Complex-Numbers.html

However, it takes the pre-C99 approach of defining its own complex types 
 and defining a set of functions to do arithmetic (so that every 
primitive arithmetic operation is a function call).  The last time it 
came up on the GSL mailing list (in 2008), Brian Gough felt it was 
important to continue to support pre-C99 compilers.

In any case updating this infrastructure in GSL seems to be a major 
undertaking, not really one that I have time to take on, unfortunately, 
even if they are interested in such an overhaul at this time.

Also, GSL has a policy of re-implementing everything in C, which doesn't 
interest me for special functions where there are perfectly good 
public-domain (or otherwise free/open-source) Fortran routines in SLATEC 
and elsewhere.  The only reason that I wrote complex erf routines was 
that we needed one ourselves (including the exponentially scaled 
varieties) and there seemed to be no usable free-software implementation.

--SGJ