This bug says zrpoly is documented but doesn't exist, which is true.
I think allroots does what zrpoly does, and uses the same algorithm.
My question is what to do about this? Leave the entry for zrpoly but
point it to allroots, or just delete the entry completely? Or maybe
add a zrpoly function that is just calls allroots?
I'm leaning towards deleting the entry completely, since it seems it
never worked anyway.
There are also other documentation bugs about non-existing functions.
I suppose that unless someone volunteers to create/recreate these,
they'll never get done. Should we just delete those entries from the
documentation?
Ray