From: dxforth@gmail.com   
      
   On 5/07/2025 6:49 pm, Anton Ertl wrote:   
   > dxf  writes:   
   > [8 stack items on the FP stack]   
   >> Puzzling because of a thread here not long ago in which scientific users   
   >> appear to suggest the opposite.  Such concerns have apparently been around   
   >> a long time:   
   >>   
   >> https://groups.google.com/g/comp.lang.forth/c/CApt6AiFkxo/m/wwZmc_Tr1PcJ   
   >   
   > I have read through the thread.  It's unclear to me which scientific   
   > users you have in mind.  My impression is that 8 stack items was   
   > deemed sufficient by many, and preferable (on 387) for efficiency   
   > reasons.   
      
   AFAICS both Skip Carter (proponent) and Julian Noble were suggesting the   
   6 level minimum were inadequate.  A similar sentiment was expressed here   
   only several months ago.  AFAIK all major forths supporting x87 hardware   
   offer software stack options.   
      
   > Certainly, of the two points this thread is about, there was a   
   > Forth200x proposal for standardizing a separate FP stack, and this   
   > proposal was accepted.  There was no proposal for increasing the   
   > minimum size of the FP stack; Forth-2012 still says:   
   >   
   > |The size of a floating-point stack shall be at least 6 items.   
      
   Only because nothing further was heard.  What became of the review   
   Elizabeth announced I've no idea.   
      
   > One interesting aspect is that VFX 5.x finally includes an FP package   
   > by default, and it started by including an SSE2-based FP package which   
   > supports a deep FP stack.  However, MPE received customer complaints   
   > about the lower number of significant digits in SSE2 (binary64)   
   > vs. 387 (80-bit FP values), so they switched the default to the   
   > 387-based FP package that only has 8 FP stack items.  Apparently no   
   > MPE customer complains about that limitation.   
   > ...   
      
   AFAIK x87 hardware stack was always MPE's main and best supported FP   
   package.  As for SSE2 it wouldn't exist if industry didn't consider   
   double-precision adequate.  My impression of MPE's SSE2 implementation   
   is that it's 'a work in progress'.  The basic precision is there but   
   transcendentals appear to be limited to single-precision.  That'd be   
   the reason I'd stick with MPE's x87 package.  Other reason is it's now   
   quite difficult and error-prone to switch FP packages as it involves   
   rebuilding the system.  The old scheme was simpler and idiot-proof.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)