Fwd: Bessel function scaling + limited range?

[John W. Eaton]

On  4-Feb-2008, Kris Kuhlman wrote:

| I have used a modified fortran90 version of these libraries, as distributed
| by Allan Miller;  TOMS 644.  The fortran code was downloaded from the
| address below.
| 
| http://users.bigpond.net.au/amiller/toms.html

OK, looking at this page, it is prefaced with

  N.B. I have been asked to provide a link to the copyright policy of
  the ACM. Loosely paraphrased, this allows use, and modification, of
  the TOMS algorithms for most non-commercial purposes. It also
  emphasizes that the ACM accepts no responsibility for the accuracy
  of the code.

so if these functions are subject to the ACM copyright, then I don't
think we can use them in Octave since the "non-commercial use only"
clause makes the license incompatible with the GPL.

OK, I see that the toms644.zip file contains a translation to Fortran
90 of an update to the original TOMS 644 algorithm.

The code in Octave comes from netlib.org, which I think is just the
original algorithm from 1986.  I assumed (perhaps incorrectly) that it
was OK to use these files, but if anyone knows differently, then we
should probably stop using them.  In any case, to avoid using code
with stupid restrictions like this, it would be good to find an
alternative.  Maybe we should just use GSL for this, as it doesn't
seem to suffer from this problem?  However, switching to GSL may not
be trivial, as it seems to rely on cblas (at least on my system) and I
think it would be best to use ATLAS and not also cblas.  If someone
would like to follow up on this, I'd suggest moving the discussion to
hte maintainers list.

Thanks,

jwe