Andreas Eder wrote:
> With the newest version of bessel.lisp I get the following errors
> with the testsuite and clisp:
>
> Running tests in rtest14:
> ********************** Problem 187 ***************
> Input:
> test_bessel(bessel_j(- 1, %i + 1.0), - .3650280288270878 %i
> - .6141603349229036, 14)
Oops. I thought I tested this with clisp, but obviously didn't.
What version of clisp are you using? I'm using 2.44.1.
In any case it appears to be a bug in clisp:
(coerce 42 'flonum) ->
*** - COERCE: 42 cannot be coerced to type (DOUBLE-FLOAT * *)
Indeed, even in clisp (no maxima loaded), this fails:
(coerce 42 '(double-float * 100d0))
(coerce 42 '(double-float 1d0))
But (coerce 42 'double-float) works.
We should report this bug to clisp, but we can probably also work around
the issue by changing our definition of flonum.
Ray