From: tkoenig@netcologne.de   
      
   Anton Ertl  schrieb:   
   > Thomas Koenig  writes:   
   >>Anton Ertl  schrieb:   
   >>   
   >>> The I32LP64 mistake   
   >>   
   >>If you consider I32LP64 a mistake, how should FORTRAN's (note the   
   >>upper case, this is pre-Fortran-90) storage association rules have   
   >>been handled, in your opinion?   
   >   
   > I am not familiar enough with FORTRAN to give a recommendation on   
   > that.  However, two observations:   
   >   
   > * The Cray-1 is primarily a Fortran machine, and it's C implementation   
   >   is ILP64, and it is successful.  So obviously an ILP64 C can live   
   >   fine with FORTRAN.   
      
   As you may know, the Cray-1 was a very special machine, which got   
   away with a lot of idiosyncracies because it was blindingly fast   
   (and caused users a lot of trouble with conversion between DOUBLE   
   PRECISION and REAL).   
      
   But that was in the late 1970s.  By the time the 64-bit worksations   
   were being designed, REAL was firmly established as 32-bit and   
   DOUBLE PRECISION as 64-bit, from the /360, the PDP-11, the VAX   
   and the very 32-bit workstations that the 64-bit workstations were   
   supposed to replace.   
      
   >   
   > * Whatever inconvenience ILP64 would have caused to Fortran   
   >   implementors is small compared to the cost in performance and   
   >   reliability that I32LP64 has cost in the C world and the cost in   
   >   encoding space (and thus code size) and implementation effort and   
   >   transistors (probably not that many, but still) that it is costing   
   >   all customers of 64-bit processors.   
      
   A 64-bit REAL and (consequently) a 128-bit DOUBLE PRECISION   
   would have made the 64-bit workstaions pretty much unusable for   
   scientific use, and a lot of these were aimed at the technical   
   and scientific market, and that meant FORTRAN.   
      
   So, put yourself into the shoes of the people designing workstations   
   RS4000 they could allow their scientific and technical customers   
   to use the same codes "as is", with no conversion, or tell them   
   they cannot use 32-bit REAL any more, and that they need to rewrite   
   all their software.   
      
   What would they have expected their customers to do?  Buy a system   
   which forces them to do this, or buy a competitor's system where   
   they can just recompile their software?   
      
   You're always harping about how compilers should be bug-comptatible   
   to previous releases.  Well, that would have been the mother of   
   all incompatiblities, aka business suicide.   
   --   
   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)