I want to make sure I have thorough coverage of exercising
the Jupiter moon calculations before I start tinkering
with truncating the L1.2 models; the eventual goal is
to decrease the code size, especially later for JavaScript.
Only 101 test cases seemed far too small. Now there are 5001.
Increased the time range to cover the years 1931..2068.
Decreased the time step from 100 days to 10 days.
Eliminated all but the geometric data, because I'm testing
code without any light-time or aberration correction.
Updated the README.txt instructions for generating the
JPL Horizons test data.
Stellarium uses the same L1.2 model for Jupiter's moons that
I am implementing. I wanted to confirm that I have valid test
data that matches something authoritative.
