On 12/22/07, Viktor T. Toth <vttoth at vttoth.com> wrote:
> 128
> 2
> is(errcatch(rat(x )) = [])
>
> Result:
> true
> The same test is successful using GCL, CLISP, and SBCL,
> the failure occurs only with CMUCL.
Well, the test is question is to test whether rat handles a bignum
exponent correctly. I guess for CMUCL, it doesn't.
The test was added recently, so it seems likely the bug was present
in earlier versions of Maxima + CMUCL but it wasn't triggered in the
test suite. Can you try rat(x^(2^128)) in an older version?
At this point I'm inclined to ignore this failure, since it is an old
bug which, after recently patching PCOEFADD, only affects one Lisp flavor.
I'd rather not delay the release any more. There will be another
release in 4 months. There are hundreds of known bugs; this one
has greater visibility since it appears in the test suite, but
it is not any more serious than others.
best
Robert