-----macrakis at gmail.com wrote: -----
>To: "Barton Willis" , maxima
>From: "Stavros Macrakis"
>Sent by: macrakis at gmail.com
>Date: 08/06/2007 03:32AM
>Subject: Re: [Maxima] float ^ rational bug in 5.12
>1) It is always wrong to give an exact (rational) result for a float
>calculation.
Agreeded---Maxima evaluates (-4.2)^(2/3) to (1.0*21^(2/3))/5^(2/3).
Notice the weird factor of 1.0.
As for the best choice for the branch, I don't know. Surely the clearest
path is
to eliminate the real-branch rule option from Maxima. That appeals to me,
but
it seems that all U.S. calculus texts assume the real-branch rule. For
powers,
Maple seems to use a consistent choice for the complex argument (-%pi,
%pi].
Does Maple (or others) have a real-branch rule option? If so, how does it
work?
Barton