Re: People are still trying to get away with disagreeing with the semantics of the x86 language

On 7/4/24 8:51 AM, olcott wrote:

On 7/4/2024 6:05 AM, joes wrote:

Am Sun, 30 Jun 2024 19:27:50 -0500 schrieb olcott:

On 6/30/2024 7:13 PM, Richard Damon wrote:

On 6/30/24 8:00 PM, olcott wrote:

>
THIS SEQUENCE CANNOT POSSIBLY RETURN WHY PERSISTENTLY LIE ABOUT IT?
>

But it does, just after H gives up its simulation.
You have even show that with a simulation.

>

   DDD correctly emulated by HHH calls an emulated HHH(DDD)
that emulates its own DDD that calls an emulated HHH(DDD)
that is either aborted at some point never returning or hits
out-of-memory error never returning

Running out of memory is only a physical constraint of no concern
to the theoretical behaviour.
>

 None-the-less it makes it totally clear that DDD correctly simulated
by HHH DOES NOT HALT.
 

No, because THIS HHH didn't run out of memery.
Your argument just shows you are trying to play a shell game.
The problem is that you want DDD to be a "template" that changes as you change HHH, but templates don't HAVE "Behavior" only the instances of them that are programs do.
I suppose you could try to work on an extension that just like "Termination Analysis" looks at the behavior of a specific program, but over all inputs, so broadens the halting problem, you could work on a Template version over all instances of the template, but you will first need to figure how to define what is a valid function to make an instance on, as any template that calls a function will become non-halting if that function is just itself non-halting, so the field becomes trivial.