Another optimisation, saves about 1% (inline one branch rather than c…

…ompute the jump)

decompressors/WBMW decompressor.asm

@@ -1,6 +1,7 @@
 WBMWDecompressTiles:  
   ld b, 4 ; bitplanes
--:push bc
+---:
+  push bc
     ; Set VRAM address
     ld c, $bf
     out (c), e
@@ -11,15 +12,30 @@ WBMWDecompressTiles:
     inc hl
     or a
     jp m, _Raw
-    jr nz, _RLE
+    jr z, +
+
+_RLE:
+    ; a = number of bytes
+    ld b, a
+    ; Get value to write
+    ld a, (hl)
+    inc hl
+-:  ; Output byte to VRAM
+    out (c), a
+    in f, (c) ; Skip 3 bytes
+    in f, (c)
+    in f, (c)
+    djnz - ; Repeat until done
+    jp -- ; Then get next byte
++:
     ; Zero means end of bitplane
     inc de ; Move to next bitplane
   pop bc
-  djnz -
+  djnz ---
   ret
   
 _Raw:  
-  ; Merker byte is the negated count -1..-128
+  ; Marker byte is the negated count -1..-128
   neg
   ld b, a
 -:; Write a byte
@@ -29,17 +45,4 @@ _Raw:
   in a, ($be)
   jp nz, - ; Repeat until done
   jp -- ; Then get next byte
-  
-_RLE:  
-  ; a = number of bytes
-  ld b, a
-  ; Get value to write
-  ld a, (hl)
-  inc hl
--:; Output byte to VRAM
-  out (c), a
-  in f, (c) ; Skip 3 bytes
-  in f, (c)
-  in f, (c)
-  djnz - ; Repeat until done
-  jp -- ; Then get next byte
+