rtest_taylor bugs



Barton Willis wrote:
> I'm getting errors for rtest_taylor tests 75,76,77,78,79,80,81,82, and 84.
> The problem seems to be that the function 'pow' is undefined. The 
> hayat.lisp
> function poly->ps calls pow. What's the story? 

I think this is my fault.  I recently used pow in ellipt.lisp, but pow 
is a macro that is defined AFTER ellipt.lisp is compiled.  Replacing pow 
with power fixes this.

We should really reorder some of the code so that macros and such are 
compiled early, before anyone uses them.

Sorry about that,

I also see that the test

taylor(jacobi_sn(inverse_jacobi_sn(x,k),k),x,0,4);

is probably no longer testing what it wants to test.  I added code so 
that jacobi_sn(inverse_jacobi_sn(x,m),m) simplifies to x.  This is 
controlled by triginverses, so setting triginverses to false will 
restore the test.

Ray