Re: V5 --- Professor Sipser --- Does Ben Bacarisse believe that Professor Sipser is wrong?

Op 26.aug.2024 om 20:14 schreef olcott:

On 8/26/2024 2:23 AM, Fred. Zwarts wrote:

Op 25.aug.2024 om 22:27 schreef olcott:

On 8/25/2024 3:15 PM, Fred. Zwarts wrote:

Op 25.aug.2024 om 21:34 schreef olcott:

On 8/25/2024 12:05 PM, Fred. Zwarts wrote:

Op 25.aug.2024 om 15:24 schreef olcott:

On 8/23/2024 4:07 PM, Ben Bacarisse wrote:

joes <noreply@example.org> writes:
>

Am Wed, 21 Aug 2024 20:55:52 -0500 schrieb olcott:

>

Professor Sipser clearly agreed that an H that does a finite simulation
of D is to predict the behavior of an unlimited simulation of D.

>
If the simulator *itself* would not abort. The H called by D is,
by construction, the same and *does* abort.

>
We don't really know what context Sipser was given.  I got in touch at
the time so do I know he had enough context to know that PO's ideas were
"wacky" and that had agreed to what he considered a "minor remark".
>
Since PO considers his words finely crafted and key to his so- called
work I think it's clear that Sipser did not take the "minor remark" he
agreed to to mean what PO takes it to mean!  My own take if that he
(Sipser) read it as a general remark about how to determine some cases,
i.e. that D names an input that H can partially simulate to determine
it's halting or otherwise.  We all know or could construct some such
cases.
>
I suspect he was tricked because PO used H and D as the names without
making it clear that D was constructed from H in the usual way (Sipser
uses H and D in at least one of his proofs).  Of course, he is clued in
enough know that, if D is indeed constructed from H like that, the
"minor remark" becomes true by being a hypothetical: if the moon is made
of cheese, the Martians can look forward to a fine fondue.  But,
personally, I think the professor is more straight talking than that,
and he simply took as a method that can work for some inputs.

>
<MIT Professor Sipser agreed to ONLY these verbatim words 10/13/2022>
     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 can abort its simulation of D and correctly report that D
     specifies a non-halting sequence of configurations.
</MIT Professor Sipser agreed to ONLY these verbatim words 10/13/2022>
>
If professor Sipser agreed to this and it only works for
some inputs then his agreement would have been incorrect.
There was an 18 message exchange prior to this agreement.
>
I do not believe that Professor Sipser made a mistake
because it still seems to be a simple tautology to me.
>

That's
the only way is could be seen as a "minor remark" with being accused of
being disingenuous.
>

Ben saw this right away and it seems that most everyone else simply lied
about it.

I don’t think you understood him.

>
I don't think PO even reads what people write.  He certainly works hard
to avoid addressing any points made to him.  I think it's true to say
that pretty much every paraphrase he attempts "X thinks ..." (usually
phrased as "so you are saying that black is white?") is garbage.
Understanding what other people say is low in his priorities since they
must be wrong anyway.
>
(I refuse to have anything more to do with PO directly after he was
unconscionably rude, but I do keep an eye out for my name in case he
continues to smear it.)
>

>
That people still disagree that a correct emulation
of N instructions of DDD according to the semantics
of the x86 language defines what a correct simulation
is still seems flat out dishonest to me.

That someone still refuses to see that skipping the last few instructions of a halting program is a violation of the semantics of the x86 language seems dishonest to me, in particular when several people pointed him to this error.
>

In the case of DDD correctly emulated by HHH this does
require HHH to emulate itself emulating DDD exactly one
time before HHH sees the repeating pattern.

>
A repeating, but not an infinite repeating pattern,

>
*D would never stop running unless aborted*
*D would never stop running unless aborted*
*D would never stop running unless aborted*
*D would never stop running unless aborted*
>
Are you just being dishonest?

>
Forget your dream of a non-aborting HHH. It does abort, so the 'unless' part makes it unnecessarily complicated. It stops running, because it aborts.
You can't have a HHH that is aborted, when it does not perform the abort itself.
Why don't you see that? Are you dishonest? It does abort and therefore is does not repeat infinitely. Then it halts. It stops running. Are you dishonest, or dreaming, or cheating?
>
>

>

 because HHH is programmed to abort and halt after a few cycles,

>
*It never has been AFTER A FEW CYCLES*
*It has always been until a specific condition is met*

>
It is coded to abort when it sees this 'specific' condition (after a few cycles) and then it halts.

>
I have corrected you on this too may times.
HALTS IS ONLY REACHING A FINAL HALT STATE
HALTS IS ONLY REACHING A FINAL HALT STATE
HALTS IS ONLY REACHING A FINAL HALT STATE
HALTS IS ONLY REACHING A FINAL HALT STATE
HALTS IS ONLY REACHING A FINAL HALT STATE

>
You don't listen. Preventing a halting program to reach its halt state by aborting the simulation does not prove that it has non-halting behaviour.
>
And by aborting the simulated HHH is prevented to reach this halt state. That does not change the fact that the simulated HHH would have detected the 'specific' condition and would have halted.

 OK I got it now.
 _DDD()
[00002172] 55         push ebp      ; housekeeping
[00002173] 8bec       mov ebp,esp   ; housekeeping
[00002175] 6872210000 push 00002172 ; push DDD
[0000217a] e853f4ffff call 000015d2 ; call HHH(DDD)
[0000217f] 83c404     add esp,+04
[00002182] 5d         pop ebp
[00002183] c3         ret
Size in bytes:(0018) [00002183]
 [Correctly emulated] is specified to mean emulated
according to the semantics of the x86 language.
 Unlike Ben you do not understand that neither DDD
[correctly emulated] by HHH nor HHH called by this DDD
[correctly emulated] by HHH can possibly return to
their caller.

You remind me of somebody who tells the same joke every 15 minutes, because he is short of memory.
I said many times that HHH cannot possibly reach the end of its own simulation, which proves that the simulation cannot possibly be correct.
You seem to forget it very easily.
HHH cannot possibly simulate itself correctly. The simulating HHH always aborts one cycle before the simulated HHH would have seen the same 'special' condition' and would have aborted as well and halt.
Stop using DDD, it is too complicated for you. Lets first look at the more simple case:
        int main() {
          return HHH(main);
        }
No DDD is needed to see that HHH cannot possibly simulate itself correctly. Here HHH halts and at the same time prints that there is an infinite recursion, which cannot be true if HHH halts.