🚀 go-pugleaf

There was one of these in the USML (Universal Simulation Mailing List)
archive, which apparently no longer exists.  I have a copy of the program,
which was called plates.c, but I have hacked on the copy I have, so I'd
rather not distribute it.  Problems I had with the program were
	1) The initial plates were polygonal (the voronoi polygons of
	a random point set), which did not look good.
	2) If I ran the simulator long enough to grind off the vertices,
	I got oceans full of so many little isands that ocean crossings
	would be trivial.
	3) Erosion was just wrong. I got one mountain that was 2 or 3 
	times the height of Everest.  Also, eroded material didn't really
	go anywhere.

Here is the comment from the beginning of the program:

/*
 * A program to simulate tectonics.
 *
 * Copyright 1988 by Mark Isaak.
 * You may distribute this however you like, as long as you don't sell it
 * and you keep this notice in it.
 *
 * Wish list:
 *      Do it on a sphere, not a square torus
 *      Give plates angular momentum
 *      Make some constants variable
 *      Horsts and grabens
 *      Let 2 plates fuse together and/or 1 plate split apart sometimes
 *      Astroblemes
 *      Take density of rock into account
 */