f2cl, cernlib, and maxima (Re: [Maxima] Release plans)

Raymond Toy writes:
 >     Ole> doubtful as to its suitability for translation and integration in
 >     Ole> maxima.
 > Why do you say that?  I just tried the converted erf function and it
 > works to full precision for the few cases I tried.

I'm under the impression that CERNLIB is not very rich in the
special-function area; I recall having resorted to TOMS for things
like complex gamma function and hypergeometric stuff (CERNLIB is the
default around here :-). Maybe I just didn't look hard enough...

CERNLIB is not just plain FORTRAN-77, at least it doesn't look like
it. The #ifdef's are there to make it run on whatever platform CERN
has used almost since 1954:-) I guess you can handle this with 
$ cpp -DCERNLIB_LINUX -DCERNLIB_UNIX -DCERNLIB_LNX -DCERNLIB_QMGLIBC

The CERNLIB functions generally depend on common "utilities" for error
reporting etc, but that might not be so much the case in mathlib. At
any rate, such "utilities" would have to be dealt with manually
because blindly translating everything could lead to a terrible bloat.

Reviewing what I have written above, I don't see any definite
show-stoppers. I guess I jumped because the word "CERNLIB" does not
give me a warm, fuzzy feeling, and going TOMS -> CERNLIB for special
functions sounded like downgrading to me. But as CERNLIB is GPL, we
can take whatever works from there and find the rest elsewhere.

 > I'm open to suggestions for other sources.  

SLATEC is in www.netlib.org and seems complete, but I couldn't find
out about the copyright/licensing policy. Actually, this is a somewhat
annoying issue with TOMS - you really have to look hard to find out it
is copyright ACM.

Emacs calc is serious business even though it might not sound like it.
It is definitively GPL but the code seems more appropriate for a
(future) arbitrary precision floating point project.

	Regards, Ole