From: muta...@gmail.com   
      
   On Tuesday, July 13, 2021 at 8:06:34 AM UTC+10, Rod Pemberton wrote:   
      
   > > > >They would've been wiser to split a 32-bit address across two   
   > > > >registers instead of using a 4-bit shift and add. That would've   
   > > > >allowed for an easier transition to 32-bit.   
   > >   
   > > [to Rod]   
   > >   
   > > Care to elaborate on that?   
   > >   
   > Instead of different x86 addressing modes for 16-bit and 32-bit   
   > instructions, the instructions could've been the same. 32-bit x86   
   > instructions are clean and orthogonal. 16-bit x86 addressing modes are   
   > a complete mess.   
      
   I'm not 100% sure I understand you, but I want a solution   
   that is constrained by the number of registers, including   
   segment registers, that were available in the 8086, plus   
   the need to support existing 16-bit applications in a memory   
   constrained environment.   
      
   If there was a lack of foresight, that's fine, but you can't   
   just say "well they should have used the full x64   
   instruction set".   
      
   A subset of x64 that supports 16-bit applications in a   
   memory-constrained (but more than 64k) environment   
   is fine.   
      
   What is that subset?   
      
   Thanks. Paul.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)