From: tkoenig@netcologne.de   
      
   John Levine  schrieb:   
      
   > Oh I forgot that using hex exponents meant there was no hidden bit, so   
   > in practice it lost three bits of precision on every operation.  There was   
   > a great deal of grumbling that people with 709x Fortran codes had to   
   > make everything double precision to keep getting reasonably good results.   
      
   Hacker's Delight phrases this as   
      
   "When IBM introduced the System/360 computer in 1964, numerical   
   analysts were horrified at the loss of precision of single-precision   
   arithmetic."   
      
   and then goes on to show how the distribution of floating point   
   values effectively reduces the precision of one quarter of floating   
   point values to 21.  (There's a name for, somebody's law, but it escapes   
   me at the moment).   
   --   
   This USENET posting was made without artificial intelligence,   
   artificial impertinence, artificial arrogance, artificial stupidity,   
   artificial flavorings or artificial colorants.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)