diff --git a/compiler/examples/cube3d-c64-float.p8 b/compiler/examples/cube3d-c64-float.p8
index d4acc1199..6b232f5d2 100644
--- a/compiler/examples/cube3d-c64-float.p8
+++ b/compiler/examples/cube3d-c64-float.p8
@@ -63,32 +63,26 @@
 
     sub draw_edges() {
 
-        sub toscreenx(float x, float z) -> byte {
-            return x/(5.0+z) * (height as float) as byte + width // 2
-        }
-
-        sub toscreeny(float y, float z) -> byte {
-            return y/(5.0+z) * (height as float) as byte + height // 2
-        }
-
         ; plot the points of the 3d cube
         ; first the points on the back, then the points on the front (painter algorithm)
 
         for ubyte i in 0 to len(xcoor)-1 {
             float rz = rotatedz[i]
             if rz >= 0.1 {
-                ubyte sx = toscreenx(rotatedx[i], rz) as ubyte
-                ubyte sy = toscreeny(rotatedy[i], rz) as ubyte
-                c64scr.setchrclr(sx, sy, 46, i+2)
+                float persp = (5.0+rz)/height
+                ubyte sx = rotatedx[i] / persp + width//2 as ubyte
+                ubyte sy = rotatedy[i] / persp + height//2 as ubyte
+                c64scr.setcc(sx, sy, 46, i+2)
             }
         }
 
         for ubyte i in 0 to len(xcoor)-1 {
             float rz = rotatedz[i]
             if rz < 0.1 {
-                ubyte sx = toscreenx(rotatedx[i], rz) as ubyte
-                ubyte sy = toscreeny(rotatedy[i], rz) as ubyte
-                c64scr.setchrclr(sx, sy, 81, i+2)
+                float persp = (5.0+rz)/height
+                ubyte sx = rotatedx[i] / persp + width//2 as ubyte
+                ubyte sy = rotatedy[i] / persp + height//2 as ubyte
+                c64scr.setcc(sx, sy, 81, i+2)
             }
         }
     }
diff --git a/compiler/examples/cube3d-c64.p8 b/compiler/examples/cube3d-c64.p8
index 765a0bad6..8982d2992 100644
--- a/compiler/examples/cube3d-c64.p8
+++ b/compiler/examples/cube3d-c64.p8
@@ -1,5 +1,4 @@
 %import c64utils
-%import c64flt
 
 ~ main {
 
@@ -74,30 +73,20 @@
         for ubyte i in 0 to len(xcoor)-1 {
             word rz = rotatedz[i]
             if rz >= 10 {
-                const float height_f = height
-                float persp = (rz as float + 180.0)/height_f
-                byte sx = rotatedx[i] as float / persp as byte + width//2
-                byte sy = rotatedy[i] as float / persp as byte + height//2
-                ; @todo switch to this once idiv_w is implemented: (and remove c64flt import)
-                ;word persp = (rz+180) // height
-                ;byte sx = rotatedx[i] // persp as byte + width//2
-                ;byte sy = rotatedy[i] // persp as byte + height//2
-                c64scr.setchrclr(sx as ubyte, sy as ubyte, 46, i+2)
+                word persp = (rz+200) // height
+                byte sx = rotatedx[i] // persp as byte + width//2
+                byte sy = rotatedy[i] // persp as byte + height//2
+                c64scr.setcc(sx as ubyte, sy as ubyte, 46, i+2)
             }
         }
 
         for ubyte i in 0 to len(xcoor)-1 {
             word rz = rotatedz[i]
             if rz < 10 {
-                const float height_f = height
-                float persp = (rz as float + 180.0)/height_f
-                byte sx = rotatedx[i] as float / persp as byte + width//2
-                byte sy = rotatedy[i] as float / persp as byte + height//2
-                ; @todo switch to this once idiv_w is implemented: (and remove c64flt import)
-                ;word persp = (rz+180) // height
-                ;byte sx = rotatedx[i] // persp as byte + width//2
-                ;byte sy = rotatedy[i] // persp as byte + height//2
-                c64scr.setchrclr(sx as ubyte, sy as ubyte, 81, i+2)
+                word persp = (rz+200) // height
+                byte sx = rotatedx[i] // persp as byte + width//2
+                byte sy = rotatedy[i] // persp as byte + height//2
+                c64scr.setcc(sx as ubyte, sy as ubyte, 81, i+2)
             }
         }
     }
diff --git a/compiler/examples/swirl-c64-float.p8 b/compiler/examples/swirl-c64-float.p8
index 78274fb5f..e5fabe0d4 100644
--- a/compiler/examples/swirl-c64-float.p8
+++ b/compiler/examples/swirl-c64-float.p8
@@ -14,23 +14,11 @@
         while true {
             float x = sin(t)
             float y = cos(t*1.1356)
-            ubyte xx=screenx(x)
-            ubyte yy=screeny(y)
-
-            ;c64.COLOR = color
-            ;c64scr.PLOT(xx,yy)
-            ;c64.CHROUT('Q')     ;  shift-q = filled circle
-            c64scr.setchrclr(xx, yy, 81, color)
-
+            ubyte xx=(x * width/2.2) + width/2.0 as ubyte
+            ubyte yy=(y * height/2.2) + height/2.0 as ubyte
+            c64scr.setcc(xx, yy, 81, color)
             t  += 0.08
             color++
         }
     }
-
-    sub screenx(float x) -> ubyte {
-        return (x * width/2.2) + width/2.0 as ubyte
-    }
-    sub screeny(float y) -> ubyte {
-        return (y * height/2.2) + height/2.0 as ubyte
-    }
 }
diff --git a/compiler/examples/swirl-c64.p8 b/compiler/examples/swirl-c64.p8
index d720a351b..da9dc6bda 100644
--- a/compiler/examples/swirl-c64.p8
+++ b/compiler/examples/swirl-c64.p8
@@ -16,7 +16,7 @@
             word y = cos8(msb(angley)) as word
             ubyte xx=msb(x*39) + 20  ; -127..127 -> 0..39
             ubyte yy=msb(y*24) + 12  ; -127..127 -> 0..24
-            c64scr.setchrclr(xx, yy, 81, color)
+            c64scr.setcc(xx, yy, 81, color)
 
             anglex+=800
             angley+=947
diff --git a/compiler/examples/test.p8 b/compiler/examples/test.p8
index a33cfcc32..df7d84097 100644
--- a/compiler/examples/test.p8
+++ b/compiler/examples/test.p8
@@ -21,32 +21,33 @@
     word w4
 
     sub start()  {
-        w1 = 120
-        w2 = -13
+        w1 = 12000
+        w2 = -333     ; -333
         w3 = w1 // w2
         c64scr.print_w(w3)
-        c64.CHROUT(':')
-        w1 = -120
-        w2 = 13
+        c64.CHROUT('\n')
+        w1 = -12000        ; -12000
+        w2 = 333
         w3 = w1 // w2
         c64scr.print_w(w3)
-        c64.CHROUT(':')
-        w2 = 13
+        c64.CHROUT('\n')
+        w2 = 4444
         w3 = w1 // w2
         c64scr.print_w(w3)
-        c64.CHROUT(':')
+        c64.CHROUT('\n')
 
-        w3 = w1 // -5
+        w3 = w1 // -5        ; -5
         c64scr.print_w(w3)
-        c64.CHROUT(':')
+        c64.CHROUT('\n')
 
-        w3 = w1 // -7
+        w3 = w1 // -7        ; -7
         c64scr.print_w(w3)
-        c64.CHROUT(':')
+        c64.CHROUT('\n')
 
-        w3 = w1 // -99
+        w1 = -w1
+        w3 = w1 // -999       ;-999
         c64scr.print_w(w3)
-        c64.CHROUT(':')
+        c64.CHROUT('\n')
 
         c64scr.print_ub(X)
         c64.CHROUT('\n')
diff --git a/prog8lib/c64utils.p8 b/prog8lib/c64utils.p8
index 25a139b99..397a9f793 100644
--- a/prog8lib/c64utils.p8
+++ b/prog8lib/c64utils.p8
@@ -1065,10 +1065,10 @@ _colorrows	.word  $d800 + range(0, 1000, 40)
 }
 
 		
-sub  setchrclr  (ubyte column, ubyte row, ubyte char, ubyte color)  {
+sub  setcc  (ubyte column, ubyte row, ubyte char, ubyte color)  {
 	; ---- set char+color at the given position on the screen
 	%asm {{
-		lda  setchrclr_row
+		lda  setcc_row
 		asl  a
 		tay
 		lda  setchr._screenrows+1,y
@@ -1077,15 +1077,15 @@ sub  setchrclr  (ubyte column, ubyte row, ubyte char, ubyte color)  {
 		sta  _colormod+2
 		lda  setchr._screenrows,y
 		clc
-		adc  setchrclr_column
+		adc  setcc_column
 		sta  _charmod+1
 		sta  _colormod+1
 		bcc  +
 		inc  _charmod+2
 		inc  _colormod+2
-+		lda  setchrclr_char
++		lda  setcc_char
 _charmod	sta  $ffff		; modified
-		lda  setchrclr_color
+		lda  setcc_color
 _colormod	sta  $ffff		; modified
 		rts
 	}}	
diff --git a/prog8lib/prog8lib.p8 b/prog8lib/prog8lib.p8
index cdbdddd6e..d6d457013 100644
--- a/prog8lib/prog8lib.p8
+++ b/prog8lib/prog8lib.p8
@@ -165,11 +165,11 @@ mul_word	.proc
 		
 idiv_b		.proc
 		; signed division: use unsigned division and fix sign of result afterwards
-		lda  c64.ESTACK_LO+1,x
-		eor  c64.ESTACK_LO+2,x
-		php			; save sign of result
 		inx
 		lda  c64.ESTACK_LO,x
+		eor  c64.ESTACK_LO+1,x
+		php			; save sign of result
+		lda  c64.ESTACK_LO,x
 		bpl  +
 		eor  #$ff
 		sec
@@ -206,7 +206,31 @@ idiv_ub		.proc
 		.pend
 		
 idiv_w		.proc
-		.error "idiv_w not yet implemented"
+		; signed division: use unsigned division and fix sign of result afterwards
+		lda  c64.ESTACK_HI+2,x
+		eor  c64.ESTACK_HI+1,x
+		php				; save sign of result
+		lda  c64.ESTACK_HI+1,x
+		bpl  +
+		jsr  neg_w			; make value positive
++		inx
+		lda  c64.ESTACK_HI+1,x
+		bpl  +
+		jsr  neg_w			; make value positive
++		lda  c64.ESTACK_LO+1,x
+		sta  c64.SCRATCH_ZPWORD1
+		lda  c64.ESTACK_HI+1,x
+		sta  c64.SCRATCH_ZPWORD1+1
+		lda  c64.ESTACK_LO,x
+		ldy  c64.ESTACK_HI,x
+		jsr  math.divmod_uw_asm
+		sta  c64.ESTACK_LO+1,x
+		tya
+		sta  c64.ESTACK_HI+1,x
+		plp
+		bpl  +
+		jmp  neg_w		; negate result
++		rts
 		.pend
 		
 idiv_uw		.proc