dos33fsprogs/demos/fire/fire_high.s
Vince Weaver b1238af49d re-arranged the entire directory structure
this will probably upset people
2021-01-05 15:29:31 -05:00

156 lines
3.9 KiB
ArmAsm

; modified to avoid using zero-page addresses
; Apple ][ Lo-res fire animation, size-optimized to 64 bytes
; by deater (Vince Weaver) <vince@deater.net>
; originally based on code described here
; http://fabiensanglard.net/doom_fire_psx/
; but this is a slightly different algorithm (but still cool looking)
; Optimization history:
; 80 bytes -- the size of tire_tiny.s
; 64 bytes -- use the firmware SCROLL routine to handle the scrolling
; this adds some flicker but saves many bytes
; 63 bytes -- qkumba noted that SCROLL leaves BAS2L:BAS2H pointing at $7d0
; 66 bytes -- tried on real hardware. Found that the original firmware
; never sets MIXCLR on boot, and my II+ machine was leaving
; it off so for correctness need to add it on.
; 66 bytes -- wasted a lot of time trying to either find clever ways to
; get $C050/C052 accessed, or to shorten the lookup table
; in half by lsr. Nothing panned out.
; 64 bytes -- realized I could save/restore the random SEEDL on the stack!
; doesn't seem to affect the output pattern either.
; Zero Page
WNDTOP = $22
WNDBTM = $23
CV = $25
BASL = $28
BASH = $29
BAS2L = $2A
BAS2H = $2B
SEEDL = $4E
; Soft Switches
SET_GR = $C050 ; Enable graphics
MIXCLR = $C052 ; Full screen, no text at bottom
LORES = $C056 ; Enable LORES graphics
; Monitor ROM routines. Try to use well-known entry points as
; these did sometimes change with newer models
SCROLL = $FC70
VTAB = $FC22 ; takes row in CV, Result is in BASL:BASH ($28/$29)
VTABZ = $FC24 ; VTABZ variant takes row in Accumulator
fire_demo:
; Set lores graphics
bit SET_GR ; force graphics mode ; 3
; LORES and LOWSCR (lo-res page1)
; are set by the boot rom
bit MIXCLR ; want full-screen graphics w/o mixed ; 3
; text the boot process doesn't set this
; (it doesn't matter in text mode)
; and while IIe and emulators come up clear
; my Apple II+ doesn't
; Set window. This seems to be set properly at boot though
; lda #0
; sta WNDTOP
; lda #24
; sta WNDBTM
scroll_loop:
jsr SCROLL ; scrolls screen up one row ; 3
;======================================
; re-draw bottom line (row 23) as white
; Y is left at 40 after SCROLL
; also BAS2L:BAS2H is left at $7d0
; this code over-writes $7F8 (the MSLOT screen hole value)
; but this should not matter for this demo
lda #$ff ; top/bottom white ; 2
w_loop:
sta (BAS2L),Y ; hline 23 (46+47) ; 2
dey ; 1
bpl w_loop ; 2
;============
; 8
fire_loop:
lda #22 ; start at line 22 ; 2
sta CV ; store in Row location ; 2
yloop:
; puts address of row CV into BASL:BASH
jsr VTAB ; 3
; loop for X values 39 ... 0
ldy #39 ; 2
xloop:
;=============================
; random8
;=============================
; 8-bit 6502 Random Number Generator
; Linear feedback shift register PRNG by White Flame
; http://codebase64.org/doku.php?id=base:small_fast_8-bit_prng
random8:
pla
; lda SEEDL ; 2
beq doEor ; 2
asl ; 1
beq noEor ; if the input was $80, skip the EOR ; 2
bcc noEor ; 2
doEor: eor #$1d ; 2
noEor: ;sta SEEDL ; 2
pha
; end inlined RNG
; Randomly either use current value, or decay by one
bmi no_change ; assume bit 7 is as random as bit 0 ; 2
; Lookup the new color in table
lda (BASL),Y ; load value ; 2
and #$7 ; mask off ; 2
tax ; 1
; lda <(color_progression),X ; 2
lda color_progression,X ; 3
sta (BASL),Y ; 2
no_change:
dey ; 1
bpl xloop ; 2
dec CV ; 2
bpl yloop ; 2
bmi scroll_loop ; 2
color_progression:
.byte $00 ; 8->0 ; 1000 -> 0000 ; needed
.byte $bb ; 9->11 ; 1001 -> 1011
.byte $00 ; 10->0 ; 1010 -> 0000 ; needed
.byte $aa ; 11->10 ; 1011 -> 1010
.byte $00 ; 12->0 ; 1100 -> 0000 ; don't care
.byte $99 ; 13->9 ; 1101 -> 1001
.byte $00 ; 14->0 ; 1110 -> 0000 ; don't care
.byte $dd ; 15->13 ; 1111 -> 1101