I've considered this in the past. Array operations can be quick, especially with aryop and friends, but if you want to draw general graphics, like, gfill/gtri/gcircle etc., It's faster to draw them with graphics commands, rather than create some functions that "draw" them into an array which is later GLOADed.
Actually, I think I created something like this for my TSCREEN thingamajig. It's a system that draws 400×240 graphics using text characters only. I think I used some internal string manipulation to get it going, but it's essentially the same concept. It's also super slow, compared to graphics.
Faster rendering using GLOAD?
DFrostCreated:
I have a proposition: set array elements linearly(2d array, elements form a line):
0,0,0,0,0,0,0,0,0,0
1,0,0,0,0,0,0,0,0,0
0,1,0,0,0,0,0,0,0,0
0,0,1,0,0,0,0,0,0,0
0,0,0,1,0,0,0,0,0,0
...
Then, use this array on GLOAD to load the frame
You could use this along with graphics commands to get processing power from both the CPU and graphics card, rather than mainly the graphics card. You could render a random scene with tris, then use GSAVE to get the frame, do some post - processing(Ie. Anti - aliasing, ), and overlap the graphics frame with the new frame made by the CPU
think about it:
what's faster?
GSPOIT(x,y) or a=array[x,y]