From: janburse@fastmail.fm   
      
   Hi,   
      
   Again, I nowhere said that (its all documented here, you   
   can re-read). I only explained that I am currently using   
   a meta-interpreter for single dispatch, I did NOT   
   say I use homoiconicity for CAS data.   
      
   I use single dispatch for CAS algorihms, but again this   
   doesn't give any implications for the internal datastructures   
   used. I have a framework that allows plugable translators   
      
   between the internal datastructure and some external   
   representation. This translation is seen by a tiny   
   detail in the Prolog toplevel. The working of this framework   
      
   is very subtle, when Prolog answer substitutions are   
   not display via the (=)/2 operator, but via the   
   (is)/2 operator it was externalized:   
      
   Here is an example:   
      
   ?- X is (A+B)^2.   
   X is A^2+2*A*B+B^2    /* small detail, result is shown   
                             with (is)/2 instead of (=)/2 */   
      
   ?- X is (A+B)^2, write(X), nl.   
   polynom(0rd4d14e06,[2-1,1-   
   polynom(0r7e813e8,[1-2]),0-   
   polynom(0r7e813e8,[2-1])])  / thats the current internal   
                                  data structure for a   
                                  multivariate polynomial */   
      
   There is really no use in explaining me CAS. I can only   
   repeat that I studied for lets say 4 years computer algebra   
   systems at ETHZ, under Prof. Erwin Engeler in the 80s.   
      
   Those times also people from Maple where at ETHZ, but   
   they came in slighly later when I already did finish   
   my studies. Others might know Roman Mäder from Mathematica.   
      
   The only thing that interests me right now are GB test cases   
   and I am looking forward to the future when I will nuke maxima   
   in that my system will beat it by lengths. I suggest to   
   make a bet, I can offeer you   
      
        A bar of chocolate   
        http://iml.jou.ufl.edu/projects/fall12/engroff_h/types.html   
      
   Or if you would like we could make another bet:   
      
        Bet between RJF and Jan Burse   
        10'000 USD (in words 10 thousend US dollars)   
      
   I already feel pitty for the old lady maxima, lets say in 3   
   years from now, maximas performance will be nowhere compared   
   to the performance of my system.   
      
   Have a Nice day!   
      
   Richard Fateman schrieb:   
   > in yet another message you seem to suggest that the essential   
   > characteristic of both lisp and prolog for computer algebra   
   > is homoiconicity   (that is, code = data).  This is false in   
   > several respects.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)