[Ecls-list] trouble building Maxima w/ ECL

[Robert Dodier]

On 6/21/10, Robert Dodier <robert.dodier at gmail.com> wrote:

> OK, thanks. I updated ECL and built Maxima successfully.
> The test suite has several failures in rtest16, several having to
> do with power series and a few with numerical evaluation of the
> zeta function.

I looked at the powerseries errors a little bit.
Looks like the problem is the treatment of special variables
in src/series.lisp, in particular VAR (I don't know if other variables
are causing trouble too). VAR is also a special variable in
src/schatc.lisp (pattern matching code). Calls to M1 in src/schatc.lisp
seem to get messed up because VAR has the wrong value.

If I load src/series.lisp into Maxima + ECL, so I guess it is interpreted,
then powerseries works as expected.

An example for which powerseries fails is powerseries(1/(1 - x)^2, x, 0).
The expected result is 'sum((i1 + 1)*x^i1, i1, 0, inf).
Maxima + ECL says "Unable to expand"; this is a result of the
failed pattern matching.

It is certainly true that Maxima's extensive use of special variables
is a disaster, but anyway the code should work as it stands.

I don;t remember if calls to powerseries worked before, or if this
is an old bug.

best,

Robert Dodier