[stella] playfield RAM usage
Subject: [stella] playfield RAM usage|
From: Glenn Saunders <cybpunks@xxxxxxxxxxxx>
Date: Sun, 04 Mar 2001 01:52:48 -0800
At 04:26 AM 3/4/2001 -0500, you wrote:
That's what I did for INV way back when, and the playfield manipulations
wound up mutating and growing and engulfing the entire kernel :) It
winds up taking a fair chunk of both RAM and kernel CPU, but it could work
pretty well here.
What's the most efficient way to store playfield in RAM anyway? I'm only
using PF1 and PF2 (hence 4 bytes per scanline) but for a full-width
playfield bitmap PF0 is only going to use a nibble.
How did Surround do it? Even using only PF1 and PF2 I'm winding up using
80 bytes for a 32 pixel by 20 8-pixel-high bitmap of blocks. If you used 6
bytes per line, that would be 120 bytes right there. So bit packing seems
to be a necessity. I guess you could store the first instance of PF0 in
the lower nibble and the second instance in the higher nibble, thus shaving
things down to 100 bytes, then somehow swapping the bits in
mid-kernel. Mind you, that's with a somewhat vertically squashed screen,
even considering score.
Imagine if Atari had gone with only 64 bytes RAM. So many games would have
128 bytes is barely enough to do playfield bitmaps.
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/