>
> By defining the principal value Log(z) in terms of
> rectform acting at the top level of the definition
> of z, we can automatically work with the
> principle value of asin(zz)
1) rectform/polarform are as naive (or more) as the rest of Maxima about
principal values.
2) Taking log of the rectform of something messy will give something even
messier. The log of the polarform is much more sensible.
3) Of course, if there are bugs in polarform, that doesn't help....
4) In general, both will be big messes, partly because Maxima is trying to
be careful by using atan2, but Maxima doesn't know how to simplify atan2
very much.
-s