From: mailbox@dmitry-kazakov.de   
      
   On Fri, 23 Apr 2010 00:47:36 GMT, Don Geddis wrote:   
      
   > "Dmitry A. Kazakov"  wrote on Sat, 17 Apr 2010:   
   >> The litmus test is. Is the behavior a property of the program, or is   
   >> it of the given description layer of?   
   >   
   > Your "litmus test" isn't even meaningful.   
      
   In what sense it is not? Either the behavior is a property of the program   
   or it is not.   
      
   > Everything is a model.   
      
   No. The word "everything" automatically renders such statements wrong. If   
   everything is a model then there is nothing it does model.   
      
   >> A separate program can modify *other* program. It does not change the   
   >> behavior of its own   
   >   
   > Who gets to define the lines, between what is program 1 and what is   
   > program 2? Why is your division necessarily the right one?   
      
   The division is arbitrary. The proposition holds by any given division.   
      
   > What is   
   > wrong with saying that the bundle together is a single program (which,   
   > overall, changes its behavior); of course one part of it is changing a   
   > different part, but that doesn't suddenly make it two separate programs.   
      
   What is wrong in pulling yourself up by your own hair?   
      
   (One part pulls another. Hey, what are you telling me about particles? Hair   
   is a higher level model to pull up!... The answer: laws of physics)   
      
   > Yes of course high level behavior depends on the correct functioning of   
   > the lower levels.   
      
   It has nothing to do with the abstraction level. It has with what is under   
   scrutiny.   
      
   >  But that wasn't what I asked.  What I asked was: how   
   > do you USE the lower level description, in order to make predictions and   
   > draw conclusions ABOUT the high level behavior?   
      
   Lover level of WHAT? Before talking about a subject it has to be fixed   
   whatever level at.   
      
   >> By no means I prefer low level models. The argument is independent on the   
   >> abstraction level of the given programming language. It is to program   
   >> correctness. How do you define it, if not based on the behavior? If   
   >> correctness is irrelevant, why do we test and debug programs?   
   >   
   > Just because you don't understand the answer to these questions, doesn't   
   > mean that there is no answer.   
      
   I know the answer, but I wished to hear your answer. Not surprisingly it   
   [stripping weasel words] is that the behavior is fundamentally undefined.   
   This was my point. (Either you have physics or Muenchhausen's stories)   
      
   >>> The state/transition model of a bubble sort algorithm is very different   
   >>> from the state/transition model of a quicksort algorithm.  At the   
   >>> algorithm level, it's an entirely different model.   
   >>   
   >> The property "sorted" of the input does not depend on the [model of]   
   >> algorithm of sorting, obviously.   
   >   
   > That's not the part that changes ... obviously.   
      
   Then your argument was wrong. Note again, the source of your confusion is   
   conflating parts of the universe of discourse, like different programs,   
   inputs and the program it processes, the program and its observer etc.   
      
   > Why are you confused to think, that if a program changes its own   
   > behavior, somehow there is nothing at all we can know about it?   
      
   Worse, it is inconsistent.   
      
   --   
   Regards,   
   Dmitry A. Kazakov   
   http://www.dmitry-kazakov.de   
      
   [ comp.ai is moderated ... your article may take a while to appear. ]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)