Apfp (Arbitrary Precision Floating Point) is a Ruby class for performing arbitrary precision floating point calculations. It also includes a class Apc for calculating with arbitrary precision complex numbers. It also contains a Real class for built-in reals. Both classes keep an estimate of the accumulated error.
omnisode generates a Ruby program to use long Taylor series to solve systems of ordinary differential equations. It generates code to solve the equations in either Maple or Maxima. Using the Taylor series, estimates are made of the location and order of poles. It has been greatly re-worked the last few years, and tested quite a bit recently. Support may be extended to other languages in the future.