Mondestrunken

Good news for the history project: I’ve (for the most part) finished implementing the Ptolemaic and Copernican models of the solar system, at least as far as longitudes go. Due to time constraints (i.e., I ought to turn it in this week), I think that’s about all I’ll be doing.

I also fixed part of the problem with my implementation of Ptolemy’s lunar model. Messing around with it again, I finally discovered a bug: I was measuring lunar longitude relative to solar longitude, but the rate of longitude change that Ptolemy specifies is relative to the first point of Aries. In other words, I had the moon moving too fast, since I was measuring speed relative to the wrong thing. Adjusting the parameter appropriately leads to a less-visibly-incorrect model; at least now you don’t have Ptolemy’s moon overtaking NASA’s moon once a year.

Ideally I’d also implement the latitudinal parts of the two models, but latitudes generally aren’t too important. Ptolemy and Copernicus both explicitly state that you don’t need to worry about latitude when computing planetary longitudes, and by definition the latitude of the sun is exactly zero. Only for the moon does it really make a difference, since you need the latitude in order to predict eclipses, but given that I need to finish the blasted thing here pretty soon I’ll wrap things up here. Besides, it’s the longitudes that are really interesting about the two models, and simulating those highlights the key difference between geocentric and heliocentric theories.

Most of the work left falls into one of the following categories:

  • Fixing a few major speed issues when running off a CD.
  • Try to correct the position-at-the-epoch values in the models when possible.
  • Simplify the user interface significantly.
  • Do a relatively small write-up explaining how to use the program, describing the models, pointing out some interesting things about them, and list the bibliography.

2 Responses

  1. Last, send a copy of your program to Wes so he can look at it. It sounds pretty cool.

  2. That shouldn’t be a problem. Except for the NASA ephemeris tables (which are about 81 MB unpreprocessed, 21 MB preprocessed), the program is fairly small.

    And it does look pretty cool.

Comments are closed.