diff --git a/tfv/NOTES b/tfv/NOTES
new file mode 100644
index 00000000..7775d738
--- /dev/null
+++ b/tfv/NOTES
@@ -0,0 +1,14 @@
+Could make VLIN faster by writing in chunks of two
+
+6502 assembly woes:
+	why do 
+		STA XX,Y
+	and
+		STA (XX),Y
+	do different things?
+
+	why are
+		LDA	$44
+	and
+		LDA	#$44
+	both valid?
diff --git a/tfv/utils.s b/tfv/utils.s
index 6df9a1a7..2b4cee34 100644
--- a/tfv/utils.s
+++ b/tfv/utils.s
@@ -421,57 +421,52 @@ done_print_string:
 	;=========================================
 	; X, V2 at Y
 vlin:
-;int vlin(int page, int y1, int y2, int at) {
 
 	sty	TEMPY		; save Y (x location)
 vlin_loop:
-;        for(i=A;i<V2;i++) {
-
 
 	txa			; a=x	(get first y)
 	and	#$fe		; Clear bottom bit
-	tay			; y=A
-	lda	gr_offsets,Y	; lookup low-res memory address
-	sta	GBASL
+	tay			;
+	lda	gr_offsets,Y	; lookup low-res memory address low
+	sta	GBASL		; put it into our indirect pointer
 	iny
-	lda	gr_offsets,Y
+	lda	gr_offsets,Y	; lookup low-res memory address high
 	clc
 	adc	DRAW_PAGE	; add in draw page offset
-	sta	GBASH
+	sta	GBASH		; put into top of indirect
 
-;               vlin_hi=i&1;
+	ldy	TEMPY		; load back in y (x offset)
 
-;	pha
+	txa			; load back in x (current y)
+	lsr			; check the low bit
+	bcc	vlin_low	; if not set, skip to low
 
-	ldy	TEMPY
+vlin_high:
+	lda	#$F0		; setup masks
+	sta	MASK
+	lda	#$0f
+	bcs	vlin_too_slow
 
-	lda	COLOR
-	sta	(GBASL),Y
+vlin_low:			; setup masks
+	lda	#$0f
+	sta	MASK
+	lda	#$f0
+vlin_too_slow:
 
-;	lda	#$0F
-;	and	(GBASL),Y
-;	sta	(GBASL),Y
-;	lda	COLOR
-;	and	#$f0
-;	ora	(GBASL),Y
-;	sta	(GBASL),Y
+	and	(GBASL),Y	; mask current byte
+	sta	(GBASL),Y	; and store back
 
- ;               if (vlin_hi) {
-  ;                      ram[vlin_addr]=ram[vlin_addr]&0x0f;
-   ;                     ram[vlin_addr]|=ram[COLOR]&0xf0;
-    ;            }
-     ;           else {
-      ;                  ram[vlin_addr]=ram[vlin_addr]&0xf0;
-       ;                 ram[vlin_addr]|=ram[COLOR]&0x0f;
-        ;        }
+	lda	MASK		; mask the color
+	and	COLOR
+	ora	(GBASL),Y	; or into the right place
+	sta	(GBASL),Y	; store it
 
-;	pla
+	inx			; increment X (current y)
+	cpx	V2		; compare to the limit
+	bcc	vlin_loop	; if <= then loop
 
-	inx
-	cpx	V2
-	bcc	vlin_loop
-
-	rts
+	rts			; return