Boing2600! update

[supercat]

- Z-gravity (i.e. ball bounces "into" the screen as though it were a top-down view). This would require scaling calculations, but all things considered it probably wouldn't be any more taxing or difficult than rotation, which is really just rotational scaling.

 

The original Boing demo, and all demos I know of on machines less powerful than the Amiga, prerendered the shape of the ball in all the rotational positions, so being able to rotate the ball is no real feat.

 

Z-scaling the ball with its oblique axis of rotation would probably be beyond the capabilities of any of those machines, but if the axis of rotation were oriented in the screen's vertical direction one could probably do a very convincing job of it. The only problem would be that the lines on the ball woudl be straight horizontal on the screen. While that would be correct if the ball were viewed from infinite distance with an infinite focal length lens, the purpose of z-scaling would be to highlight the fact that the camera was close enough that vertical motion would affect size.

 

The memory requirements to pull off the trick would probably be fairly modest. I would expect it should be possible to render a ball up scalable from 12x20 to 48x80 pixels in real time, with eight rotational "steps", in a 4K cart on the 2600. Who wants me to try?