Interesting life program

[Paul Callahan]

d...@pollux.acs.oakland.edu (Daniel Finster) writes:

>Even though this is a good 8 years old now, it is still the finest Life
>simulation I have seen on the MS-DOS platform. It is extremely fast, and
>supports a lot of interesting ways to modify the simulation.

After testing this program, I can only think of one advantage it has
over Alan Hensel's most recent release, which is that it allows the
definition of arbitrary 3x3 neighborhood transition rules, whereas
Hensel's is limited to totalistic rules (those based on cell counts).
However, this is enough of an advantage to keep it around.

It is reasonably fast (though considerably slower than Hensel's
program), and achieves speed by exploiting pattern sparsity and
probably some bit parallelism. It has the major disadvantage of
fixing the total grid size to that which fits on the screen, and a
user interface that is insufficient for pattern development, as it
does not allow cut and paste or even (to my knowledge) the positioning
of loaded patterns. The pattern save format is unfortunately not
human readable, and I have not been able to determine how to convert
between it and Xlife or Hensel's life. If anyone wishes to work on
this, I'm interested in such a conversion program.

...getting back to the one advantage:

The rule editor made it possible to set up the Ship rule variation of
life, which is the minimum deviation from standard Life rules that
supports the replicator of HighLife (or S23/B36 Life, discussed in a
recent article by David Bell). In the Ship rule, almost all
neighborhoods behave as in Life, with the exception of the two
neighborhoods:

**. .**
*.* and *.*
.** **.

In the case of these "ship" neighborhoods, birth will occur at the center
cell, thus deviating from the "overcrowding" rule of Life (HighLife allows
such a birth in all neighborhoods containing 6 cells). To save effort for
anyone who wishes to experiment with the replicator under the ship rule,
I have enclosed a uuencoded .zip file of both the rules and the replicator
pattern:

--
Paul Callahan
call...@biffvm.cs.jhu.edu