On Thu, 2008-03-13 at 08:17 -0400, Raymond Toy (RT/EUS) wrote:
> Bob Baker wrote:
> > Well, I tried running Maxima directly in a command window. I gave the
> > commands and got exactly the same reply. So I don't think Euler can be
> > blamed for it. Did you try these commands to see what happens on your
> > system?
>
> Yes, I did. That's how I knew it was asking a question. Doesn't it ask
> you? Did it run for a long time before the error message? It runs for
> quite a while, but I stopped it after a bit.
>
> What version are you using?
Using 5.14.0 + Clisp, I get the same result as Raymond (a question and
an integration that never ends). Using 5.14.0 + GCL, I get the same
error message Bob is getting and no questions.
Bob, notice that your integral could be written in a simpler form:
f(x):=log(2-cos(x)+sqrt(3+cos(x)*(cos(x)-4)))$
integrate(2*(1-exp(-abs(1)*f(x))*cos(3*x))/(exp(f(x))-exp(-f(x))),x,0,%
pi);
(I don't know how to make Maxima expand sinh(x) in terms of
exponentials, without doing the same for sin(x)).
using that simpler form, 5.14.0+clisp immediately gives up, while 5.14.0
+GCL now gives a different error message:
Too many contexts.
-- an error. To debug this try debugmode(true);
Hope that helps someone to tells us what GCL is doing differently from
Clisp.
Regards,
Jaime Villate