From: dxforth@gmail.com   
      
   On 10/07/2025 2:35 pm, Paul Rubin wrote:   
   > dxf  writes:   
   >> I suspect IEEE simply standardized what had become common practice among   
   >> implementers.   
   >   
   > No, it was really new and interesting.   
   > https://people.eecs.berkeley.edu/~wkahan/ieee754status/754story.html   
   >   
   >> What little I know about SSE2 it's not as well thought out or organized   
   >> as Intel's original effort.  E.g. doing something as simple as changing   
   >> sign of an fp number is a pain when NANs are factored in.   
   >   
   > I wonder if later SSE/AVX/whatever versions fixed this stuff.   
      
   Actually I was wrong.  x87 FCHS (aka FNEGATE) changes the sign bit of a NAN.   
   IEEE doesn't consider NANs to be signed even though x87 implementations may   
   display them that way.  The catch with SSE is there's nothing like FCHS or FABS   
   so depending on how one implements them, results vary across implementations.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)