Re: DDD correctly emulated by H0

On 6/25/2024 10:41 AM, Mike Terry wrote:

On 25/06/2024 14:46, Alan Mackenzie wrote:

[ Followup-To: set ]
>
In comp.theory olcott <polcott333@gmail.com> wrote:

On 6/25/2024 4:22 AM, joes wrote:

Am Sat, 22 Jun 2024 13:47:24 -0500 schrieb olcott:

On 6/22/2024 1:39 PM, Fred. Zwarts wrote:

Op 21.jun.2024 om 15:21 schreef olcott:

>

When we stipulate that the only measure of a correct emulation is the
semantics of the x86 programming language then we see that when DDD is
correctly emulated by H0 that its call to H0(DDD) cannot possibly
return.

Yes. Which is wrong, because H0 should terminate.

>
[ .... ]
>

The call from DDD to H0(DDD) when DDD is correctly emulated
by H0 cannot possibly return.

>

Until you acknowledge this is true, this is the
only thing that I am willing to talk to you about.

>
I think you are talking at cross purposes.  Joes's point is that H0
should terminate because it's a decider.  You're saying that when H0 is
"correctly" emulating, it won't terminate.  I don't recall seeing anybody
arguing against that.

 Hehe, everyone has an opinion on what PO is saying! :)
 So here's mine:  I think PO is saying that when DDD is correctly *emulated* [not "correctly /emulating/"] then "it" will not return.
 To be clear, that's saying that the /emulation/ does not get as far as the final return, i.e. outer H0 will stop emulating (aka "abort") without having emulated DDD's return statement.  PO has acknowledged that the outer H0 will return after aborting its emulation.
 So what? I hear everyone asking.  So what indeed! The "behaviour" of a partial emulation is not the behaviour of the computation itself [trivially] and not what the HP is about. But PO then gets hopelessly muddled, thinking at least two wrong things:
 

_DDD()
[00002172] 55               push ebp      ; housekeeping
[00002173] 8bec             mov ebp,esp   ; housekeeping
[00002175] 6872210000       push 00002172 ; push DDD
[0000217a] e853f4ffff       call 000015d2 ; call H0(DDD)
[0000217f] 83c404           add esp,+04
[00002182] 5d               pop ebp
[00002183] c3               ret
Size in bytes:(0018) [00002183]
The call from DDD to H0(DDD) (when DDD is correctly emulated
by any H0 that can possibly exist at machine address 0000217a)
H0 cannot possibly return.
*In other words you insist on flatly disagreeing with the*
*semantics of the x86 programming language* Not very smart of you.

1)  That DDD would /never/ have terminated in "1-oo steps of emulation", i.e. that DDD really doesn't halt, not simply that H0 aborted it before it returned.  (Thats muddling all the different examples of H in his head each with their own personalised (D,D) inputs into one single super-H/super-D.  Of course, the FIXED DDD under discussion /does/ in fact return in some finite number of steps.  H0 is also fixed and is coded in a way that aborts /before/ that number of steps is emulated.)
 

Strawman deception. DDD correctly simulated by any H0 that can possibly
exist at machine address 0000217a is not the same as DDD correctly
simulated by any H1 that DDD never calls.
The strawman deception is a kind of lie. Why lie?

2)  PO knows that the /reason/ H0 decided to abort was that it matched his "infinite recursive emulation" pattern in the emulation trace - therefore he believes it's correct to decide non-halting because his pattern proves that.  (That's just Wrong, but PO really really really believes the pattern is sound, so that's the end of it.  He has no intention or capability of ever attempting to prove his rule is sound.)
  Mike.
 

--
Copyright 2024 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer