Am Freitag, den 04.06.2010, 09:59 -0600 schrieb Robert Dodier:
> On 5/25/10, Stavros Macrakis <macrakis at alum.mit.edu> wrote:
>
> > If I'm not mistaken, this all goes back to the fact that simpabs calls
> > cabs. There is a reason that abs and cabs are distinct -- one is a
> > mathematical (simplifying) function, the other is a form-changing function.
>
> It seems like it would help to rename cabs to something more
> descriptive, then. How about expand_abs ?
I think we should not change the name. cabs is good enough.
> > simpabs should never call cabs. (This bug was apparently introduced some
> > time before Maxima got into sourceforge; it was not in MIT Macsyma when I
> > was working on it.)
>
> OK by me, although I don't have a strong opinion about this.
> It would be interesting to disable any calls from simpabs to cabs
> just to see what would happen.
Today, I have tried it. I have cut out the calls to cabs completely, but
I have introduced code to simplify only complex numbers. As expected I
have got a lot of problems with the testsuite. One problem is that a lot
of expected simplifications like abs(exp(%i*x)) -> 1 are done by cabs
and have to be reimplemented into abs.
Dieter Kaiser