7 replies to this topic

[tschak909]

So far the kernel seems to be coming along very nicely, I was finally able to implement a smooth scrolling algorithm by unrolling everything.... but I have some other problems that have come up:

* The first digit for the score is garbled, not sure why.. it's looking like a mangled pointer somehow....
* The scroll completes, then starts at the very beginning again. I suspect this is because i am updating state during overscan.. do I need to move the bounds enforcement (basically the AND), in-kernel?

I am pasting source and a screenshot.. but I'm showing some progress!



and the current code:
 salmonrun_070311.zip   36.88K   50 downloads

Until next time!

-Thom

[SeaGtGruff]

tschak909, on Sun Mar 11, 2007 2:57 AM, said:

* The first digit for the score is garbled, not sure why.. it's looking like a mangled pointer somehow....

Yes, this is where it gets mangled:

SCOREPTR5		= $CC; ...
SCANCOUNTER3		= $CD; and another counter...

The one-byte counter SCANCOUNTER3 is overlapping the two-byte pointer SCOREPTR5.

Michael

[SeaGtGruff]

This is also looks wrong:

PF1DATA			= $81; 32 bytes of PF1.
PF2DATA			= $A0; 32 bytes of PF2.
SCANCOUNTER		= $C0; a scanline counter.

If PF1DATA is supposed to be 32 bytes, then PF2DATA should start at $A1, and SCANCOUNTER should start at $C1, with everything after that getting shifted 1 byte as well.

Michael

[tschak909]

ahhh! thanks, yeah.. that would do it!

[Andrew Davie]

tschak909, on Sun Mar 11, 2007 5:57 PM, said:

So far the kernel seems to be coming along very nicely, I was finally able to implement a smooth scrolling algorithm by unrolling everything.... but I have some other problems that have come up:

* The first digit for the score is garbled, not sure why.. it's looking like a mangled pointer somehow....
* The scroll completes, then starts at the very beginning again. I suspect this is because i am updating state during overscan.. do I need to move the bounds enforcement (basically the AND), in-kernel?

I am pasting source and a screenshot.. but I'm showing some progress!

attachment

and the current code:
attachment

Until next time!

-Thom

To avoid this sort of variable clashing, I suggest you try the following...

				SEG.U variables
				ORG $80

; define your variables here
; this is an uninitialised segment -- the assembler will calculate addresses, but create no actual data in the binary

; examples...
GAME_STATUS		ds 1		 ; 1 byte for this var!
PF1DATA			ds 32	   ; 32 for this one
PF2DATA			ds 32	   ; and this

; this allows you to insert/remove variables at will, and resize, without having to recalculate!
; in other words, let the assembler calculate the locations of stuff, you just state how big they are!
  
			SEG code
			ORG $F000

 ; rest of your program here

[SeaGtGruff]

Excellent idea, Andrew! I need to start doing that, too.

Michael

[gambler172]

Hi Thom
The old Salmon Run on the VCS?
Greeaatt
Cannot wait to see a playable bin.
greetings gambler172

[tschak909]

Yes indeed, i was always a huge fan of that game (and ALL of Bill Williams' stuff!), and I want to do a solid port (I think we can even improve on the graphics a bit since it is very vertically oriented!)

I'm progressing very quickly on this, quite amazing considering this is my first game attempt on the VCS!