From: nowhere@never.at   
      
   Kerr Mudd-John wrote:   
   ...   
   >>> Lots of fixups in there; well done on using the jmp space!   
   >>I also tried to replace the two nops right after the CRLF with:   
   >> dec bx |dec bx                                    ;,   
   >> and then adjust already 'slid' [bx+..] modifications for the second   
   >> line followed by pop bx | push bx.   
   >> It works too, but it isn't any shorter.   
      
   > You inspired me to think of a horrible^w brilliant scheme, but this margin is   
   > too small to contain it.   
      
   :)   
      
   Meanwhile I checked on another 2:1 algo which uses only one single sub   
   (no "+/=") and a first "0" is treated different (jb iterate| jz special).   
   It drastically limits the available instruction range [no 4B..B6,D1..D6].   
   So this "0"-2nd -constant is required to create at least AA and AC.   
   I think the 4:3 decoder can be done that way (even w/o 80..83).   
   __   
   wolfgang   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)