diff --git a/games/keen/story.s b/games/keen/story.s
index e529a8ee..6c33b858 100644
--- a/games/keen/story.s
+++ b/games/keen/story.s
@@ -57,89 +57,172 @@ load_background:
 	bit	SET_TEXT
 	bit	PAGE1
 
-
 	lda	#<story_data
 	sta	START_LINE_L
 	lda	#>story_data
 	sta	START_LINE_H
 
+	;===========================
+	; main loop
+	;===========================
+	; 16 + 16 * (16 + 40*(39) +7) + 17 + 17*40
+	; 16 + 16*(1583) + 17 + 680
+	;	16 + 25328 + 17 + 680
+	; 26041 = 400 lines :(  we're on 16*8=128
+
 redraw_text:
 
-	ldx	#0
+	ldx	#0							; 2
 
-	lda	START_LINE_L
-	sta	INL
-	lda	START_LINE_H
-	sta	INH
+	lda	START_LINE_L						; 3
+	sta	input_smc+1						; 4
+	lda	START_LINE_H						; 3
+	sta	input_smc+2						; 4
+								;===========
+								;	16
 
 outer_text_loop:
 
-	lda	gr_offsets_low,X
-	sta	OUTL
-	lda	gr_offsets_high,X
-	sta	OUTH
+	lda	gr_offsets_low,X					; 4+
+	sta	OUTL							; 3
+	lda	gr_offsets_high,X					; 4+
+	sta	OUTH							; 3
 
+	ldy	#0							; 2
+								;============
+								;	16
 
-	ldy	#39
 inner_text_loop:
-	lda	(INL),Y
-	sta	(OUTL),Y
-	dey
-	bpl	inner_text_loop
 
-	clc
-	lda	INL
-	adc	#40
-	sta	INL
-	lda	INH
-	adc	#0
-	sta	INH
+input_smc:
+	lda	story_bg						; 4
+	sta	(OUTL),Y						; 6
 
-	inx
-	cpx	#17
-	bne	outer_text_loop
+	clc								; 2
+	lda	input_smc+1						; 4
+	adc	#1							; 2
+	sta	input_smc+1						; 4
+	lda	input_smc+2						; 4
+	adc	#0							; 2
+	sta	input_smc+2						; 4
 
+	iny								; 2
+	cpy	#40							; 2
+	bne	inner_text_loop						; 2/3
+								;============
+								;	39
+
+									; -1
+
+
+	inx								; 2
+	cpx	#16							; 2
+	bne	outer_text_loop						; 2/3
+								;============
+								;	7
+
+									; -1
 	;==================
 	; draw message
 	;==================
 
-	ldx	#18
-	lda	gr_offsets_low,X
-	sta	OUTL
-	lda	gr_offsets_high,X
-	sta	OUTH
+	ldx	#17							; 2
+	lda	gr_offsets_low,X					; 4+
+	sta	OUTL							; 3
+	lda	gr_offsets_high,X					; 4+
+	sta	OUTH							; 3
+
+	ldy	#39							; 2
+							;==================
+							;		17
 
-	ldy	#39
 message_text_loop:
-	lda	message,Y
-	and	#$3f
-	sta	(OUTL),Y
-	dey
-	bpl	message_text_loop
+	lda	message,Y						; 4+
+	and	#$3f							; 2
+	sta	(OUTL),Y						; 6
+	dey								; 2
+	bpl	message_text_loop					; 2/3
 
+							;==================
+							;		17*40
 
-	jsr	wait_until_keypress
+							;		-1
 
-	and	#$7f		; clear high bit
-	and	#$df		; change lower to upper
+	;===================================
+	; can we do this constant time?
+	;===================================
 
-	cmp	#13
-	beq	done_with_story
-	cmp	#27
-	beq	done_with_story
+check_keypress:
+	lda	KEYPRESS						; 4
+	bpl	done_key7						; 2/3
+; 6
+	bit	KEYRESET						; 4
 
-	cmp	#'W'
-	beq	do_up
-	cmp	#$0B
-	beq	do_up
+	and	#$7f		; clear high bit			; 2
+	and	#$df		; change lower to upper			; 2
+; 14
+	cmp	#13							; 2
+	beq	done_with_story						; 2/3
+; 18
+	cmp	#27							; 2
+	beq	done_with_story						; 2/3
+; 22
+	cmp	#'W'							; 2
+	beq	do_up							; 2/3
+; 26
+	cmp	#$0B							; 2
+	beq	do_up							; 2/3
+; 30
+	cmp	#'S'							; 2
+	beq	do_down							; 2/3
+; 34
+	cmp	#$0A							; 2
+	beq	do_down							; 2/3
+; 38
+	bne	done_key41		; bra				; 3
 
-	cmp	#'S'
-	beq	do_down
-	cmp	#$0A
-	beq	do_down
+done_key7:
+	; need to waste 34 cycles
+
+	inc	$00	; 5
+	inc	$00	; 5
+	inc	$00	; 5
+	inc	$00	; 5
+	inc	$00	; 5
+	inc	$00	; 5
+	nop
+	nop
+done_key41:
+	inc	$00
+	nop
+
+	; 18*8*65= 9360
+	; want to delay 9360 - 41 - 7 - 4 = 9308 - 20 = 9288/9 = 1032
+	;	1032/256= 4 r 8
+	;
+
+	lda	#4							; 2
+	ldy	#8							; 2
+	jsr	delay_loop
+
+	; want to delay 6*8*65 = 3120+4550 = 7670
+	; want to delay 7670 - 15 - 12 = 7643 - 20 = 7623/9  = 847
+	; 905/256=3 r 79
+
+	inc	$00	; nop5
+	inc	$00	; nop5
+	nop		; nop2
+
+	bit	SET_GR							; 4
+
+	lda	#3							; 2
+	ldy	#79							; 2
+	jsr	delay_loop
+	bit	SET_TEXT						; 4
 
 done_key:
-	jmp	redraw_text
+
+	jmp	check_keypress						; 3
 
 do_up:
 	lda	START_LINE_H
@@ -171,7 +254,6 @@ do_down:
 	beq	done_key
 
 down_ok:
-
 	clc
 	lda	START_LINE_L
 	adc	#40
@@ -181,7 +263,6 @@ down_ok:
 	sta	START_LINE_H
 	jmp	redraw_text
 
-
 done_with_story:
 
 
@@ -199,7 +280,7 @@ done_with_story:
 	.include	"gr_copy.s"
 ;	.include	"wait_a_bit.s"
 	.include	"gr_offsets.s"
-	.include	"gr_offsets_split.s"
+
 	.include	"zx02_optim.s"
 
 	.include	"gr_fast_clear.s"
@@ -222,6 +303,27 @@ wait_until_keypress:
 	bit	KEYRESET
 	rts
 
+.align $100
 
 message:
 	.byte "      ESC TO EXIT / ARROWS TO READ      ",0
+
+	.include	"gr_offsets_split.s"
+
+
+	;=====================================
+	; short delay by Bruce Clark
+	;   any delay between 8 to 589832 with res of 9
+	;=====================================
+	; 9*(256*A+Y)+8 + 12 for jsr/rts
+	; A and Y both $FF at the end
+
+size_delay:
+
+delay_loop:
+	cpy	#1
+	dey
+	sbc	#0
+	bcs	delay_loop
+delay_12:
+	rts