tests with maxima 5.5 beta3



I just began to run my favorite tests for 
computer algebra systems in maxima 5.5 beta3, after
having tried: 
(a+b+c+d+1)^35
on my linux box
(amd k6-2 at 300 MHz, 256 MB ram, 
512 MB swap, Mandrake 7.1): 21 seconds.

(By the way, 
1. Maple V.5.1 cannot do the job:
   Error, (in expand/bigpow)
   object too large
      3,869
2. MuPAD 2.0 succeeds in 401s after
   _at least_ one reinitialization of the kernel
   (doubling the stack for pari?)
3. pari/gp 2.1 succeeds in 97s with a big initial stack
)

The problem:
------------
I am very confused by the result of
integrate(1/(1+x^5+x),x);
which is, in my opinion, false and very strange, without
any warning.

Best regards.
----
Daniel Duparc <daniel.duparc@libertysurf.fr>
29 av. de la Commune de Paris
94400 Vitry sur Seine (France)