Nick S Bensema wrote:
> >The physics are kinda secondary, and I'd settle for any at all. The
> >real thing is switching from hard-coded horizon/distance events to
> >elegant equations for everything.
>
> Math is a bitch and a half on the 6502 to begin with, but on the 2600
> it has to be fast. How about tables? Defender uses tables to multiply
> by 40 (though I haven't ascertained exactly why it needs to do that
> so often, yet).
That's kinda how it works now. In a very simple form (straight flat
track) it works great. The track is defined by both horizontal and
vertical relative adjustments, and tables determine how each object will
look in position relative to the horizon (x and z axis) and the current
point of view (y axis). But that isn't the way it's supposed to work.
That's just a place marker engine to perfect graphics and gameplay
ideas. The problem of course with tables to determine so much is
re-doing them whem you decide a couple extra pixels or scan lines would
look pretty (or make the controls feel better)... Small adjustments
become big projects. House of cards. Reducing the problem down to just
a couple simple tables that efficiently perform some workhorse chores
and are not too hard to adjust is probably where I'm headed.
> I'm probably the only human being in a 600 mile radius who understands
> anything about this particular topic. Certainly without the list, or
> at least Usenet, I would have seen little reason to continue.
Or even get started ;)
--
mor@xxxxxxx
http://www.crl.com/~mor/
--
To unsubscribe, send the word UNSUBSCRIBE in the body of a message to
stella-request@xxxxxxxxxxx
