Hi Christopher!
While I'm trying to assimilate all you've written
(thanks a zillion!) let me explain some thing, I'm not
sure you're aware of.
> This will eliminate the whole "CMP #$20/BMI
> NoMeteorSprite" testing
I've a data structure for things the kernel can display.
The kernel just executes and draws these objects,
without knowing what is what, they all are treated
equal, obviously I don't have the time in the kernel for
distinguishing between drawing a ship or a meteor or
anything.
Now, this is the structure:
xpos ds #MAXOBJECTS ; horizontal position
ypos ds #MAXOBJECTS ; vertical position
xymov ds #MAXOBJECTS ; xy motion
zmov ds #MAXOBJECTS ; z motion
sprtsizetype ds #MAXOBJECTS ; size/type/color
(Ten objects = 50 bytes!)
To let the kernel draw these, I'm constantly Y sorting
them.
That means in the first frame the motherships data might
for example be stored in
xpos+3
ypos+3
xymov+3
zmov+3
sprtsizetype+3
in the next frame it might be located in
xpos+7
ypos+7
xymov+7
zmov+7
sprtsizetype+7
So, I originally don't know in the object handler which
object is which and I'm forced to test this, no other
choice. Except sacrificing another 10 bytes RAM for
indexing the object types, but I'm not that desperate -
yet :-)
But thanks for all you other insights.
Again, I originally had this big chunky mess of
straightforward *grown* code. It was when I started
loosing control over it, that I decided to clearly
extract it into subs with proper defined functionality.
It was good to do that anyway, since now I've clearly
defined subs, I can optimize them all separately for
speed and having all functionality seperated in logic
blocks helps me mentally entwining the chaos.
In future steps I'll now carefully go back step by step
to the "big mess", only it'll be way less chaotic and
more optimized then :-)
Greetings,
Manuel
----------------------------------------------------------------------------------------------
Archives (includes files) at http://www.biglist.com/lists/stella/archives/
Unsub & more at http://www.biglist.com/lists/stella/
