ABOUT OOLIFE

This project was done in response to comments made on comp.object
concerning the speed of an object implementation of Conway's life.

In particular, one comment was that Life might be too simple a
hack to be worth an object implementation.  I agree with Gall
who said "A complex system that works is invariably found to have
evolved from a simple system that worked" and therefore I believe
that anything worth coding is worth doing in an OO language.  On
the occassion that we program something that never grows beyond
a two hundred line kludge, we lose little having tread the object
path.  But if that kludge grows into a serious application, then
we have a foundation that will be much easier to extend.

With that in mind, I put together this code to prove that even in
a CPU bound process like Life, the loss from object languages is
minimal.

What I found is that the life processing allowed me to run 6
generations a second on a 10,000 cell colony under Windows/NT.

WARNING: When running under DOS you will have memory problems.  This
is due to the 640k barrier.  The development machine had 620k free
base and supported under 2,500 cells in the simulation.  While the
cells are portable the DOS test.cpp code contains Borland specific
graphics code which prevents it from running anywhere but DOS.

Rich Wellner
5/10/94