Strange thing in Maxima: reduction of fraction



On THU, 7 DEC 2006, Raymond Toy wrote:

>>>>>> "Leo" == Leo  <sdl.web at gmail.com> writes:
>
>     Leo> On THU, 7 DEC 2006, Raymond Toy wrote:
>     Leo> [...]
>
>     Leo> Looks like a bug.
>     >> 
>     >> Is it?  Maxima doesn't always simplify everything that you type in.
>     >> Isn't %o2 correct, independent of any assumptions on p & q (except
>     >> that I think maxima is implicitly assuming p and q are real)?
>     >> 
>     >> Ray
>
>     Leo> But what about %o1 output? Isn't that a bug?
>
> Why?  If p and q are real, isn't abs(p-q)^2 always equal to (p-q)^2?
> Or do you mean that %o1 didn't simplify to something else?
>
> Ray
>

I just don't understand why the returns are so different.

(q-p)/abs(q-p)^2 => 1/(q-p)
(p-q)/abs(p-q)^2 => (p-q)/(q-p)^2

-- 
Leo