integrate(log(1+(1+x**2)**0.5), x) -> incorrect result



Stavros Macrakis wrote:
> On Fri, Nov 28, 2008 at 10:10 PM, Raymond Toy <toy.raymond at gmail.com
> <mailto:toy.raymond at gmail.com>> wrote:
>
>     >   integrate(log(1+(1+x**2)**(1/2)), x)
>
>     Maxima can't compute the integral completely by itself, unfortunately,
>     but it does get close:
>
>
FWIW, the original integral with 0.5 instead of 1/2 fails in rischint. 
In both cases we get to the point of trying to integrate
x^2/(1+x^2+(1+x^2)^(1/2)).  When the power is 1/2, rischint returns the
noun form.  But when the power is 0.5, rischint returns the incorrect
answer.

I don't understand the rischint code, so I don't know why it produces
the wrong answer.

One obvious workaround is to rationalize the expression before calling
rischint so that it never has to deal with float numbers.

Ray