...\author

This work was supported in part by NSF Grant number CCR-9214963 and by NSF Infrastructure Grant number CDA-8722788. and DIGITAL LIBRARY grant XXX

...

Why did we program in Lisp? Several reasons:

1. We like Lisp, especially for exploratory programming.
2. Natural data-structures are linked lists.
3. Convenient built-in package (``Common Windows'') for bit-map display. Interaction with the program for learning and debugging have been substantially assisted by the easy availability of this set of routines providing access to the X-window interface. This is not essential for the running of the core routines, but only for checking on results.
4. (After some work) convenient access to files as formatted from the scanners.
5. Lisp programs tend to be quite portable: Portability to other implementations of Common Lisp would require a translation of the foreign-function call scheme in the Allegro dialect to some other form.

.

...

Our initial code required a constant step size, but we found this too restrictive. Our current program allows for fractions: a step size of one degree is not exactly 57 to 1, but a sequence: (57 58 57 57 57 57 58 ...).

...

If we were using pixel arrays, we would expect the processing time to decrease by a factor of four. In own representation, while the number of rows would be halved; we would not expect the number of intervals per row to halve as well. Intervals would be lost only when details were lost to the decreased resolution.

...page.)

Actually, David Glowacki.

...page.)

provided by David Glowacki, too