Re: 197 page execution trace of DDD correctly simulated by HHH

On 7/1/2024 9:27 AM, Fred. Zwarts wrote:

Op 01.jul.2024 om 14:57 schreef olcott:

On 7/1/2024 3:27 AM, Fred. Zwarts wrote:

Op 30.jun.2024 om 19:25 schreef olcott:

On 6/30/2024 3:42 AM, joes wrote:
>

No, I mean: why does the inner simulator repeat instead of aborting,
the same as the outer one does?
>

Technically it is called detecting a repeating state.

Yeah, I know. My point is: all recursive calls both enter and detect
a repeating state.
>

>
The inner ones always see one less execution trace
than the next outer one, thus could only meet their
abort criteria after they have already been aborted.
>

>
Which indicates that they were aborted too soon, showing that the emulation was incorrect.

>
Unless the outer HHH aborts its simulation after some
fixed number of correct emulations or none of the HHH
ever aborts and HHH never stops running.

>
But that does not make the result of the abort correct.

 

Not aborting will loop infinitely.

      If simulating halt decider H correctly simulates its input D
     until H correctly determines that its simulated D would never
     stop running unless aborted
 THEN H IS NECESSARILY CORRECT TO ABORT
THEN H IS NECESSARILY CORRECT TO ABORT
THEN H IS NECESSARILY CORRECT TO ABORT

 

Aborting will abort too soon. Both cases are incorrect.
>

 It doesn't seem like you care about the truth.
 

>
There is no passing the guy in front of you if you both
continue to run at the exact same speed, he will always
be ahead of you.
>

>
But shooting his tyres is also incorrect.