david sharp wrote:

> In 2d there are the classic 'Conway' rules. So far as I know, there is no
> standard set of rules
> for 3d. With certain rule sets you can get static formations and gliders and
> blinkers, etc.
> So it seems pretty wide open for experimenting.
> But a problem with my POV implementation is that once the 'playing field'
> gets large, the
> computation time gets unacceptable if the goal is experimenting, like
> looking for forms that
> 'do stuff'.

Damn.  Yet another cpu bounded situation.  What we need is a massive cluster 
of computers all rendering frames.  But we can't because the GOL frame depends 
on the previous frame.  Each system would need to render a few lines only and 
then assemble the image and then move onto the next frame.

https://www.blastwave.com/solaris/parts/SS10SS20U1/Image02.jpg

I have a stack of systems that just sit around and collect dust.  I've always 
wanted to get sixteen of them going simutaneously but the scripting would be a 
bit nasty and the assembly of the image even nastier.

Just thinking and typeing at the same time.

Dennis

Post a reply to this message