On 1/6/24 3:54 PM, olcott wrote:
> On 1/6/24 21:54, olcott wrote:
>> On 1/6/2024 2:14 PM, immibis wrote:
>>> On 1/6/24 20:20, olcott wrote:
>>>> On 1/6/2024 12:53 PM, Richard Damon wrote:
>>>>> On 1/6/24 1:32 PM, olcott wrote:
>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>
>>>>>>>> Whether or not a question is incorrect is always determined by ...
>>>>>>>
>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>
>>>>>>
>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>> mistake.
>>>>>>
>>>>>
>>>>> Then you need to actually show that the question is incorrect.
>>>>>
>>>>
>>>> Incorrect questions are defined as any question that cannot possibly
>>>> have a correct answer.
>>>>
>>>> This remains the case even when the question is only incorrect when
>>>> posed to a specific entity within the discourse context, thus Jack's
>>>> question posed to Jack and input D to H <are> incorrect questions.
>>>>
>>>> That logicians and computer scientists are ignorant of these details
>>>> it no excuse for ignoring them.
>>>>
>>>
>>> Correct question: "Does D(D) halt?"
>>>
>>> Incorrect question: "Given that H is a correct decider for the
>>> halting problem, does D(D) halt?"
>>>
>>
>> I will use your approach to this:
>> Encode the context directly in the question.
>>
>> "Does a barber who shaves every man who does not shave himself shave
>> himself?" has no correct answer.
>>
>> Thus the question is not "Does D(D) halt?"
>>
>> The question (with full context added) is:
>>
>> "What correct Boolean value does H return when D is defined to do the
>> opposite of whatever value that H returns?"
>>
>>
> No, that isn't the question.
>
> The question is "Is there an H, such that for all P and I, H(P,I)
> returns true if and only if P(I) halts?"
>
> And the answer is "no, there is no such H."
>
> The question is "Is there a barber who shaves all men who do not shave
> themselves?"
>
> And the answer is "no, there is no such barber."

No, that CAN'T be the question, as H can only return 1 answer, the
answer that its algoritm generates. Rmmember, H is a SPECIFIC program,
and D is a SPECIFIC input (built on a specific decider).

The Question is, and only is, "Does the Machine/Input represented by the
input Halt when run?"

Any variation is just a strawman, which by your own words, your using
makes you dishonest.

H is correct, if its answer agrees with the correct answer to that question.

You are a category error in your logic, showing you are totally ignorant
about COmputation Theory.

You are just showing you don't understand how logic works.

On 1/6/24 7:17 PM, olcott wrote:
> On 1/6/2024 2:58 PM, immibis wrote:
>> On 1/6/24 21:54, olcott wrote:
>>> On 1/6/2024 2:14 PM, immibis wrote:
>>>> On 1/6/24 20:20, olcott wrote:
>>>>> On 1/6/2024 12:53 PM, Richard Damon wrote:
>>>>>> On 1/6/24 1:32 PM, olcott wrote:
>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>
>>>>>>>>> Whether or not a question is incorrect is always determined by ...
>>>>>>>>
>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>
>>>>>>>
>>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>>> mistake.
>>>>>>>
>>>>>>
>>>>>> Then you need to actually show that the question is incorrect.
>>>>>>
>>>>>
>>>>> Incorrect questions are defined as any question that cannot possibly
>>>>> have a correct answer.
>>>>>
>>>>> This remains the case even when the question is only incorrect when
>>>>> posed to a specific entity within the discourse context, thus Jack's
>>>>> question posed to Jack and input D to H <are> incorrect questions.
>>>>>
>>>>> That logicians and computer scientists are ignorant of these details
>>>>> it no excuse for ignoring them.
>>>>>
>>>>
>>>> Correct question: "Does D(D) halt?"
>>>>
>>>> Incorrect question: "Given that H is a correct decider for the
>>>> halting problem, does D(D) halt?"
>>>>
>>>
>>> I will use your approach to this:
>>> Encode the context directly in the question.
>>>
>>> "Does a barber who shaves every man who does not shave himself shave
>>> himself?" has no correct answer.
>>>
>>> Thus the question is not "Does D(D) halt?"
>>>
>>> The question (with full context added) is:
>>>
>>> "What correct Boolean value does H return when D is defined to do the
>>> opposite of whatever value that H returns?"
>>>
>>>
>> No, that isn't the question.
>
> I used the exact same method that you used on the barber paradox by
> encoding the full context of the question directly in the question.
> *Please do not be dishonest*
>

No, you didn't, because he didn't.

"Does the Barber shave himself?", is a perfectly valid question that can
be asked about ANY actual Barber, and always has a correct answer (when
asked about any actual Barber)

Just as, "does the machine described by the input Halt when run?" can be
asked of any decider (as long as the input actually represents a machine).

The "invalid" part of the question, if you want to find a problem, is
that the Barber doesn't actually exist, because its definition is invalid.

If you want to find the same problem with the input, then you need to
say the input doesn't actually exist, but D does exist, and is a valid
machine, as long as the H it is built on exists and is a valid machine.

So, your claim is that Your H that you claim is a correct decider, must
not actually exist or is not a valid machine, so it can't actually be
the "Correct Decider" you claim it to be.

So, you are just admitting that you logic is broken,

On 1/6/24 3:48 PM, olcott wrote:
> On 1/6/2024 2:11 PM, immibis wrote:
>> On 1/6/24 19:32, olcott wrote:
>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>> On 1/6/24 17:07, olcott wrote:
>>>>>
>>>>> Whether or not a question is incorrect is always determined by ...
>>>>
>>>> What is the purpose of talking about incorrect questions?
>>>>
>>>
>>> Proving that the whole notion of undecidability is anchored in a
>>> mistake.
>>>
>> What mistake?
>
> Undecidability is anchored in the fact that incorrect questions
> have no correct answer. The correct approach is to reject these
> incorrect questions and it is a mistake to not reject them.
>

Nope.

Undecidability is anchored in the fact that some questions with correct
answer need an infinite number of steps in that system to determine that
answer.

You are just proving you don't understand what you are talking about.

The questions HAVE correct answers.

On 1/6/24 7:24 PM, olcott wrote:
> On 1/6/2024 2:58 PM, immibis wrote:
>> On 1/6/24 21:48, olcott wrote:
>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>> On 1/6/24 19:32, olcott wrote:
>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>
>>>>>>> Whether or not a question is incorrect is always determined by ...
>>>>>>
>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>
>>>>>
>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>> mistake.
>>>>>
>>>> What mistake?
>>>
>>> Undecidability is anchored in the fact that incorrect questions
>>> have no correct answer. The correct approach is to reject these
>>> incorrect questions and it is a mistake to not reject them.
>>>
>>
>> Undecidability is actually the nonexistence of a Turing machine which
>> always gives the correct answer to a certain question. That is the
>> definition.
>
> When we screen out every question that cannot possibly have a
> correct true/false return value as invalid input then undecidability
> cannot possibly exist because these are the undecidable inputs.
>

THen why can't we make a Halt Decider?

All specific inputs have a correct answer.

You are just stuck in your insanity that you think specific inouts given
to specific machines are exactly the same thing as sets of input s given
to sets of machines, and that different things can be the same.

In other words, you are basing your logic on lies.

On 1/7/24 05:19, Richard Damon wrote:
> On 1/6/24 3:48 PM, olcott wrote:
>> On 1/6/2024 2:11 PM, immibis wrote:
>>> On 1/6/24 19:32, olcott wrote:
>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>
>>>>>> Whether or not a question is incorrect is always determined by ...
>>>>>
>>>>> What is the purpose of talking about incorrect questions?
>>>>>
>>>>
>>>> Proving that the whole notion of undecidability is anchored in a
>>>> mistake.
>>>>
>>> What mistake?
>>
>> Undecidability is anchored in the fact that incorrect questions
>> have no correct answer. The correct approach is to reject these
>> incorrect questions and it is a mistake to not reject them.
>>
>
> Nope.
>
> Undecidability is anchored in the fact that some questions with correct
> answer need an infinite number of steps in that system to determine that
> answer.
>

Nope.

You cannot take an infinite number of steps. An answer is not determined
at all if it "needs" an infinite number of steps.

On 1/7/24 4:19 PM, immibis wrote:
> On 1/7/24 05:19, Richard Damon wrote:
>> On 1/6/24 3:48 PM, olcott wrote:
>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>> On 1/6/24 19:32, olcott wrote:
>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>
>>>>>>> Whether or not a question is incorrect is always determined by ...
>>>>>>
>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>
>>>>>
>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>> mistake.
>>>>>
>>>> What mistake?
>>>
>>> Undecidability is anchored in the fact that incorrect questions
>>> have no correct answer. The correct approach is to reject these
>>> incorrect questions and it is a mistake to not reject them.
>>>
>>
>> Nope.
>>
>> Undecidability is anchored in the fact that some questions with
>> correct answer need an infinite number of steps in that system to
>> determine that answer.
>>
>
> Nope.
>
> You cannot take an infinite number of steps. An answer is not determined
> at all if it "needs" an infinite number of steps.
>

That a machine does not halt, might need an infinte number of steps of
simulation to prove.

The FACT that it doesn't halt, is still established, even if we can not
"compute" that answer by simulation.

That is the definition of "Undecidability", that no Turing Machine (or
equivalent) can be created that can compute the answer.

On 1/7/24 22:57, Richard Damon wrote:
> On 1/7/24 4:19 PM, immibis wrote:
>> You cannot take an infinite number of steps. An answer is not
>> determined at all if it "needs" an infinite number of steps.
>>
>
> That a machine does not halt, might need an infinte number of steps of
> simulation to prove.

What you mean to say more precisely is that the simulation will not
prove it after any finite number of steps.

> The FACT that it doesn't halt, is still established, even if we can not
> "compute" that answer by simulation.

If a machine does not halt after any finite number of steps, then it
does not halt. This is not the same as saying it halts after an infinite
number of steps.

> That is the definition of "Undecidability", that no Turing Machine (or
> equivalent) can be created that can compute the answer.

On 1/7/2024 4:19 PM, immibis wrote:
> On 1/7/24 05:19, Richard Damon wrote:
>> On 1/6/24 3:48 PM, olcott wrote:
>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>> On 1/6/24 19:32, olcott wrote:
>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>> On 1/6/24 17:07, olcott wrote:

>>>>>>> Whether or not a question is incorrect
>>>>>>> is always determined by ...
>>>>>>
>>>>>> What is the purpose of talking about
>>>>>> incorrect questions?
>>>>>
>>>>> Proving that
>>>>> the whole notion of undecidability
>>>>> is anchored in a mistake.
>>>>
>>>> What mistake?
>>>
>>> Undecidability is anchored in the fact that
>>> incorrect questions have no correct answer.
>>> The correct approach is
>>> to reject these incorrect questions and
>>> it is a mistake to not reject them.
>>
>> Nope.
>> Undecidability is anchored in the fact
>> that some questions with correct answer need
>> an infinite number of steps in that system
>> to determine that answer.
>
> Nope.
> You cannot take an infinite number of steps.

However,
we can say "This is what a step is"
and know it to be true of
each one of infinitely.many steps.

Also,
we can know something _now_ which
_we didn't know before_ to be true of
each one of infinitely.many steps.

We can know that by making only claims which
we can see are each not.first.false.

Only finitely.many not.first.false claims.
We can avoid making infinitely.many claims,
I think.

We know that,
in a finite sequence of claims,
all claims not.first.false forces
all claims not.false.

And, in some instances,
we can _see_ merely by _looking_
that a claim is not.first.false.

For example,
suppose we make these claims, in this order:
⟨ ¬P∨Q P Q ⟩

Q is not.first.false.
We can't say very much more about Q
but we can see that Q is not.first.false.
⟨ ¬P∨Q P Q ⟩
    t  t t
    t  f t
    f  t f
    t  f f

And so on.

If we can see that all the claims we make
are, like Q, not.first.false,
then we know that all the claims we make
are true.

> You cannot take an infinite number of steps.

We cannot do that, but
we can learn about infinitely.many steps
without taking infinitely.many steps.

> An answer is not determined at all if
> it "needs" an infinite number of steps.

About "needs":
There is more than one way to skin a cat.

On 1/6/2024 9:33 PM, immibis wrote:
> On 1/7/24 04:20, olcott wrote:
>> On 1/6/2024 7:23 PM, immibis wrote:
>>> On 1/7/24 01:24, olcott wrote:
>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>> On 1/6/24 21:48, olcott wrote:
>>>>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>>>>> On 1/6/24 19:32, olcott wrote:
>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>
>>>>>>>>>> Whether or not a question is incorrect is always determined by
>>>>>>>>>> ...
>>>>>>>>>
>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>
>>>>>>>>
>>>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>>>> mistake.
>>>>>>>>
>>>>>>> What mistake?
>>>>>>
>>>>>> Undecidability is anchored in the fact that incorrect questions
>>>>>> have no correct answer. The correct approach is to reject these
>>>>>> incorrect questions and it is a mistake to not reject them.
>>>>>>
>>>>>
>>>>> Undecidability is actually the nonexistence of a Turing machine
>>>>> which always gives the correct answer to a certain question. That
>>>>> is the definition.
>>>>
>>>> When we screen out every question that cannot possibly have a
>>>> correct true/false return value as invalid input then undecidability
>>>> cannot possibly exist because these are the undecidable inputs.
>>>>
>>> The halting problem has a correct true/false return value
>>
>> *Not when the full context of the question is considered*
>> "What correct Boolean value does H return when D is defined
>>   to do the opposite of whatever value that H returns?"
>>
> That isn't the halting problem, that's a different problem.

*That <is> the full context of the question that H is asked*
H1 is asked: Does the direct execution of D(D) halt?

The conclusive proof that this IS NOT the same question
that H is asked is the easily verified fact that the fully
operational H1 gets the correct answer.

https://github.com/plolcott/x86utm/blob/master/Halt7.c
H1 is on lines 635-657
H is on lines 668-691

> The halting problem asks whether D halts.
>

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/8/24 03:12, olcott wrote:
> On 1/6/2024 9:33 PM, immibis wrote:
>> On 1/7/24 04:20, olcott wrote:
>>> On 1/6/2024 7:23 PM, immibis wrote:
>>>> On 1/7/24 01:24, olcott wrote:
>>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>>> On 1/6/24 21:48, olcott wrote:
>>>>>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>>>>>> On 1/6/24 19:32, olcott wrote:
>>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>>
>>>>>>>>>>> Whether or not a question is incorrect is always determined
>>>>>>>>>>> by ...
>>>>>>>>>>
>>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>>>>> mistake.
>>>>>>>>>
>>>>>>>> What mistake?
>>>>>>>
>>>>>>> Undecidability is anchored in the fact that incorrect questions
>>>>>>> have no correct answer. The correct approach is to reject these
>>>>>>> incorrect questions and it is a mistake to not reject them.
>>>>>>>
>>>>>>
>>>>>> Undecidability is actually the nonexistence of a Turing machine
>>>>>> which always gives the correct answer to a certain question. That
>>>>>> is the definition.
>>>>>
>>>>> When we screen out every question that cannot possibly have a
>>>>> correct true/false return value as invalid input then undecidability
>>>>> cannot possibly exist because these are the undecidable inputs.
>>>>>
>>>> The halting problem has a correct true/false return value
>>>
>>> *Not when the full context of the question is considered*
>>> "What correct Boolean value does H return when D is defined
>>>   to do the opposite of whatever value that H returns?"
>>>
>> That isn't the halting problem, that's a different problem.
>
> *That <is> the full context of the question that H is asked*

We are talking about the halting problem, not H. The halting problem is:
Does D(D) halt?

> this IS NOT the same question that H is asked

Well how do we ask H this question?

If we can't ask H whether D(D) halts, then H doesn't solve the halting
problem.

On 1/6/2024 9:40 PM, immibis wrote:
> On 1/7/24 03:49, olcott wrote:
>> On 1/6/2024 7:19 PM, immibis wrote:
>>> On 1/7/24 02:12, olcott wrote:
>>>> On 1/6/2024 6:22 PM, immibis wrote:
>>>>> On 1/7/24 01:17, olcott wrote:
>>>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>>>> On 1/6/24 21:54, olcott wrote:
>>>>>>>> On 1/6/2024 2:14 PM, immibis wrote:
>>>>>>>>> On 1/6/24 20:20, olcott wrote:
>>>>>>>>>> On 1/6/2024 12:53 PM, Richard Damon wrote:
>>>>>>>>>>> On 1/6/24 1:32 PM, olcott wrote:
>>>>>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> Whether or not a question is incorrect is always
>>>>>>>>>>>>>> determined by ...
>>>>>>>>>>>>>
>>>>>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Proving that the whole notion of undecidability is anchored
>>>>>>>>>>>> in a
>>>>>>>>>>>> mistake.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Then you need to actually show that the question is incorrect.
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Incorrect questions are defined as any question that cannot
>>>>>>>>>> possibly
>>>>>>>>>> have a correct answer.
>>>>>>>>>>
>>>>>>>>>> This remains the case even when the question is only incorrect
>>>>>>>>>> when
>>>>>>>>>> posed to a specific entity within the discourse context, thus
>>>>>>>>>> Jack's
>>>>>>>>>> question posed to Jack and input D to H <are> incorrect
>>>>>>>>>> questions.
>>>>>>>>>>
>>>>>>>>>> That logicians and computer scientists are ignorant of these
>>>>>>>>>> details
>>>>>>>>>> it no excuse for ignoring them.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Correct question: "Does D(D) halt?"
>>>>>>>>>
>>>>>>>>> Incorrect question: "Given that H is a correct decider for the
>>>>>>>>> halting problem, does D(D) halt?"
>>>>>>>>>
>>>>>>>>
>>>>>>>> I will use your approach to this:
>>>>>>>> Encode the context directly in the question.
>>>>>>>>
>>>>>>>> "Does a barber who shaves every man who does not shave himself
>>>>>>>> shave himself?" has no correct answer.
>>>>>>>>
>>>>>>>> Thus the question is not "Does D(D) halt?"
>>>>>>>>
>>>>>>>> The question (with full context added) is:
>>>>>>>>
>>>>>>>> "What correct Boolean value does H return when D is defined to
>>>>>>>> do the
>>>>>>>> opposite of whatever value that H returns?"
>>>>>>>>
>>>>>>>>
>>>>>>> No, that isn't the question.
>>>>>>
>>>>>> I used the exact same method that you used on the barber paradox by
>>>>>> encoding the full context of the question directly in the question.
>>>>>> *Please do not be dishonest*
>>>>>>
>>>>>
>>>>> It's still the wrong question.
>>>>>
>>>>
>>>> I proved otherwise by your own example, why lie?
>>>>
>>> You didn't prove anything. You used the wrong question.
>>
>> In other words you are saying:
>> Does the barber shave himself? <is> the impossible question
>> without the context details.
>
> There is a correct answer to the question in every universe.
>
> There is no universe where the context makes sense.
>
> The context is incorrect, not the question.

On 1/6/2024 1:54 PM, immibis wrote:
> "Does the barber shave himself?" has a correct answer: it's one of
> "yes" or "no".
>
> "Does a barber who shaves every man who does not shave himself shave
> himself?" has no correct answer.

*That was a brilliant insight*

*By that EXACT same reasoning*

Does the direct execution of D(D) halt? Has a correct answer: it is yes.

What correct Boolean value does H return when D is defined to do the
opposite of whatever value that H returns?" has no correct answer.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/6/2024 9:41 PM, immibis wrote:
> On 1/7/24 03:55, olcott wrote:
>> On 1/6/2024 7:19 PM, immibis wrote:
>>> On 1/7/24 02:10, olcott wrote:
>>>> On 1/6/2024 3:02 PM, immibis wrote:
>>>>> On 1/6/24 21:45, olcott wrote:
>>>>>> On 1/6/2024 1:54 PM, immibis wrote:
>>>>>>
>>>>>> There is no halt decider that works correctly on incorrect inputs.
>>>>>>
>>>>>
>>>>> There is no an incorrect input to a halt decider, except for input
>>>>> which isn't a Turing machine. If a program has incorrect input,
>>>>> that program isn't a halt decider.
>>>>
>>>> That <is> the received view, yet the received view never bothered
>>>> to think these things all-the-way through. Only when this received view
>>>> understands that the HP counter-example is isomorphic to determining
>>>> whether the Liar Paradox is true or false.
>>>>
>>>
>>> A halt decider takes any program as input, or any program and its
>>> initial tape contents (it doesn't matter which definition is used).
>>> Are you trying to redefine the word "program"?
>>
>> I have been doing primary research on pathological self-reference
>> for twenty years. It <is> the case that pathological self-reference
>> <is> an error that must be rejected. The Liar Paradox <is not> a
>> truth-bearer. Is <is> the case that all expressions that are isomorphic
>> to the Liar Paradox <are> erroneous.
>>
>
> A halt decider takes any program as input. This is a definition. You
> cannot refute a definition.

On 1/6/2024 1:54 PM, immibis wrote:
> "Does the barber shave himself?" has a correct answer: it's one of
> "yes" or "no".
>
> "Does a barber who shaves every man who does not shave himself shave
> himself?" has no correct answer.

*That was a brilliant insight*

*By that EXACT same reasoning*

Does the direct execution of D(D) halt? Has a correct answer: it is yes.

What correct Boolean value does H return when D is defined to do the
opposite of whatever value that H returns?" has no correct answer.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/7/24 9:45 PM, olcott wrote:
> On 1/6/2024 9:41 PM, immibis wrote:
>> On 1/7/24 03:55, olcott wrote:
>>> On 1/6/2024 7:19 PM, immibis wrote:
>>>> On 1/7/24 02:10, olcott wrote:
>>>>> On 1/6/2024 3:02 PM, immibis wrote:
>>>>>> On 1/6/24 21:45, olcott wrote:
>>>>>>> On 1/6/2024 1:54 PM, immibis wrote:
>>>>>>>
>>>>>>> There is no halt decider that works correctly on incorrect inputs.
>>>>>>>
>>>>>>
>>>>>> There is no an incorrect input to a halt decider, except for input
>>>>>> which isn't a Turing machine. If a program has incorrect input,
>>>>>> that program isn't a halt decider.
>>>>>
>>>>> That <is> the received view, yet the received view never bothered
>>>>> to think these things all-the-way through. Only when this received
>>>>> view
>>>>> understands that the HP counter-example is isomorphic to determining
>>>>> whether the Liar Paradox is true or false.
>>>>>
>>>>
>>>> A halt decider takes any program as input, or any program and its
>>>> initial tape contents (it doesn't matter which definition is used).
>>>> Are you trying to redefine the word "program"?
>>>
>>> I have been doing primary research on pathological self-reference
>>> for twenty years. It <is> the case that pathological self-reference
>>> <is> an error that must be rejected. The Liar Paradox <is not> a
>>> truth-bearer. Is <is> the case that all expressions that are isomorphic
>>> to the Liar Paradox <are> erroneous.
>>>
>>
>> A halt decider takes any program as input. This is a definition. You
>> cannot refute a definition.
>
> On 1/6/2024 1:54 PM, immibis wrote:
> > "Does the barber shave himself?" has a correct answer: it's one of
> > "yes" or "no".
> >
> > "Does a barber who shaves every man who does not shave himself shave
> > himself?" has no correct answer.
>
> *That was a brilliant insight*
>
> *By that EXACT same reasoning*
>
> Does the direct execution of D(D) halt? Has a correct answer: it is yes.
>
> What correct Boolean value does H return when D is defined to do the
> opposite of whatever value that H returns?" has no correct answer.
>

And, you undestand that the problem with the question:

"Does a barber who shaves every man who does ot shave himself, shave
himself?" doesn't show a problem with the question "Does Barber X shave
himself?" but with specifying a Barber who shave everyone who doesn't
shave themselves.

In the same way, asking about the boolean value that a correct H returns
when asked if D(D), where D(D) is defined to do the opposite of what H
does?" isn't about the asking of the behavior of a specific D(D), as
that always has an answer, but of the assumption that an H can exist
that returns a correct answer to this question.

You could TRY to say the error is in the relationship, but Turing Theory
shows that we can ALLWAYS build a D and give it input D, as long as the
H it is defined to be based on exists. Thus again, the error isn't in
the asking about the Halting of a Machine D given input D, but on the
assumption that your H actually exist.

Thus, all you hsve done is proved that you can't ask the question that
assumes that H can return a correct answer, because such an H doesn't exist.

In other words, you have just shown that your last twenty years were
wasted on working on an incorrect definitiom of the problem.

On 1/7/24 9:40 PM, olcott wrote:
> On 1/6/2024 9:40 PM, immibis wrote:
>> On 1/7/24 03:49, olcott wrote:
>>> On 1/6/2024 7:19 PM, immibis wrote:
>>>> On 1/7/24 02:12, olcott wrote:
>>>>> On 1/6/2024 6:22 PM, immibis wrote:
>>>>>> On 1/7/24 01:17, olcott wrote:
>>>>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>>>>> On 1/6/24 21:54, olcott wrote:
>>>>>>>>> On 1/6/2024 2:14 PM, immibis wrote:
>>>>>>>>>> On 1/6/24 20:20, olcott wrote:
>>>>>>>>>>> On 1/6/2024 12:53 PM, Richard Damon wrote:
>>>>>>>>>>>> On 1/6/24 1:32 PM, olcott wrote:
>>>>>>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>>>>>>
>>>>>>>>>>>>>>> Whether or not a question is incorrect is always
>>>>>>>>>>>>>>> determined by ...
>>>>>>>>>>>>>>
>>>>>>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>>>>>>
>>>>>>>>>>>>>
>>>>>>>>>>>>> Proving that the whole notion of undecidability is anchored
>>>>>>>>>>>>> in a
>>>>>>>>>>>>> mistake.
>>>>>>>>>>>>>
>>>>>>>>>>>>
>>>>>>>>>>>> Then you need to actually show that the question is incorrect.
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Incorrect questions are defined as any question that cannot
>>>>>>>>>>> possibly
>>>>>>>>>>> have a correct answer.
>>>>>>>>>>>
>>>>>>>>>>> This remains the case even when the question is only
>>>>>>>>>>> incorrect when
>>>>>>>>>>> posed to a specific entity within the discourse context, thus
>>>>>>>>>>> Jack's
>>>>>>>>>>> question posed to Jack and input D to H <are> incorrect
>>>>>>>>>>> questions.
>>>>>>>>>>>
>>>>>>>>>>> That logicians and computer scientists are ignorant of these
>>>>>>>>>>> details
>>>>>>>>>>> it no excuse for ignoring them.
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Correct question: "Does D(D) halt?"
>>>>>>>>>>
>>>>>>>>>> Incorrect question: "Given that H is a correct decider for the
>>>>>>>>>> halting problem, does D(D) halt?"
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> I will use your approach to this:
>>>>>>>>> Encode the context directly in the question.
>>>>>>>>>
>>>>>>>>> "Does a barber who shaves every man who does not shave himself
>>>>>>>>> shave himself?" has no correct answer.
>>>>>>>>>
>>>>>>>>> Thus the question is not "Does D(D) halt?"
>>>>>>>>>
>>>>>>>>> The question (with full context added) is:
>>>>>>>>>
>>>>>>>>> "What correct Boolean value does H return when D is defined to
>>>>>>>>> do the
>>>>>>>>> opposite of whatever value that H returns?"
>>>>>>>>>
>>>>>>>>>
>>>>>>>> No, that isn't the question.
>>>>>>>
>>>>>>> I used the exact same method that you used on the barber paradox by
>>>>>>> encoding the full context of the question directly in the question.
>>>>>>> *Please do not be dishonest*
>>>>>>>
>>>>>>
>>>>>> It's still the wrong question.
>>>>>>
>>>>>
>>>>> I proved otherwise by your own example, why lie?
>>>>>
>>>> You didn't prove anything. You used the wrong question.
>>>
>>> In other words you are saying:
>>> Does the barber shave himself? <is> the impossible question
>>> without the context details.
>>
>> There is a correct answer to the question in every universe.
>>
>> There is no universe where the context makes sense.
>>
>> The context is incorrect, not the question.
>
> On 1/6/2024 1:54 PM, immibis wrote:
> > "Does the barber shave himself?" has a correct answer: it's one of
> > "yes" or "no".
> >
> > "Does a barber who shaves every man who does not shave himself shave
> > himself?" has no correct answer.
>
> *That was a brilliant insight*
>
> *By that EXACT same reasoning*
>
> Does the direct execution of D(D) halt? Has a correct answer: it is yes.
>
> What correct Boolean value does H return when D is defined to do the
> opposite of whatever value that H returns?" has no correct answer.
>
>
>

Which means that the presumption that H CAN return a correct answer is
flawed.

Thus, the Halting Question is fine, It is the assumption that Correct
Halt Deciders exist that is flawed, which is EXACTLY what the Halting
Theorem says, so you are just proving the thing that you were trying to
disprove.

Must be awful to realize you have wasted so much time on such a simple
error on your part.

On 1/7/24 9:12 PM, olcott wrote:
> On 1/6/2024 9:33 PM, immibis wrote:
>> On 1/7/24 04:20, olcott wrote:
>>> On 1/6/2024 7:23 PM, immibis wrote:
>>>> On 1/7/24 01:24, olcott wrote:
>>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>>> On 1/6/24 21:48, olcott wrote:
>>>>>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>>>>>> On 1/6/24 19:32, olcott wrote:
>>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>>
>>>>>>>>>>> Whether or not a question is incorrect is always determined
>>>>>>>>>>> by ...
>>>>>>>>>>
>>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>>>>> mistake.
>>>>>>>>>
>>>>>>>> What mistake?
>>>>>>>
>>>>>>> Undecidability is anchored in the fact that incorrect questions
>>>>>>> have no correct answer. The correct approach is to reject these
>>>>>>> incorrect questions and it is a mistake to not reject them.
>>>>>>>
>>>>>>
>>>>>> Undecidability is actually the nonexistence of a Turing machine
>>>>>> which always gives the correct answer to a certain question. That
>>>>>> is the definition.
>>>>>
>>>>> When we screen out every question that cannot possibly have a
>>>>> correct true/false return value as invalid input then undecidability
>>>>> cannot possibly exist because these are the undecidable inputs.
>>>>>
>>>> The halting problem has a correct true/false return value
>>>
>>> *Not when the full context of the question is considered*
>>> "What correct Boolean value does H return when D is defined
>>>   to do the opposite of whatever value that H returns?"
>>>
>> That isn't the halting problem, that's a different problem.
>
> *That <is> the full context of the question that H is asked*
> H1 is asked: Does the direct execution of D(D) halt?
>
> The conclusive proof that this IS NOT the same question
> that H is asked is the easily verified fact that the fully
> operational H1 gets the correct answer.
>
> https://github.com/plolcott/x86utm/blob/master/Halt7.c
> H1 is on lines 635-657
>  H is on lines 668-691
>
>
>> The halting problem asks whether D halts.
>>
>

Nope.

That IS that halting problem.

Your "added context" is the problem, with the assumption that H itself
can return a correct answer.

H1 getting the right answer show that right answer exists.

That H doesn't give it proves H is just wrong, and you are too stupid to
understand that.

On 1/6/2024 9:42 PM, immibis wrote:
> On 1/7/24 03:57, olcott wrote:
>> On 1/6/2024 7:21 PM, immibis wrote:
>>> On 1/7/24 01:45, olcott wrote:
>>>  >
>>>> The ZFC axiom schema of specification forbids the
>>>> creation of self-referential sets.
>>>>
>>>
>>> Yes, but you missed the forest for the trees. ZFC says that
>>> self-referential sets do not exist.
>>
>> Not quite. ZFC prevents self-referential sets from coming into
>> existence. Pathological self-reference (Olcott 2004) must be
>> rejected as erroneous.
>>
>
> Is it more obvious to say: Self-referential sets contradict the ZFC
> axioms, so they can't exist in any world where the ZFC axioms are true.

It is more important to say that every undecidable problem that is
undecidable because of pathological self-reference is simply incorrect.

>>> Olcott machines say that self-referential programs do not exist. I
>>> don't understand what you think is a self-referential program, but we
>>> know how to construct one by following the steps of Linz (you said
>>> this makes a self-referential program). So what happens when we
>>> follow the steps? Is the result not a program?
>>
>
> And I suppose self-referential programs contradict the Olcott axioms, so
> they can't exist in any world where the Olcott axioms are true. The
> problem is that I just wrote one, so obviously the Olcott axioms aren't
> true in this world I live in.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/6/2024 9:43 PM, immibis wrote:
> On 1/7/24 04:00, olcott wrote:
>> On 1/6/2024 7:21 PM, immibis wrote:
>>> On 1/7/24 01:28, olcott wrote:
>>>> On 1/6/2024 3:00 PM, immibis wrote:
>>>>> On 1/6/24 21:32, olcott wrote:
>>>>>>
>>>>>> In that case embedded_H can recognize the its input is a copy of its
>>>>>> own TM specification.
>>>>>>
>>>>>
>>>>> The specification of embedded_H can be modified without altering
>>>>> the semantics. It will give the same answer as before, but it will
>>>>> no longer be a copy.
>>>>
>>>> There are several ways to recognize pathological self-reference.
>>>> I presenting the simple one that has fully operational code.
>>>> I could not encode the other one because the x86 emulator could
>>>> not handle it. None-the-less it still does work on the Peter Linz
>>>> Proof.
>>>>
>>> Your simple one isn't isomorphic to a Turing machine.
>>
>> It demonstrates a basic principle COMPLETELY without the gaps
>> in reasoning that must be included in every HP proof.
>>
>> I show how the exact same basic principle <is> applied to the
>> Peter Linz proof.
>>
>
> Your simple on isn't isomorphic to a Turing machine, because of the
> self-call-detection which has no analogue in a Turing machine.

That in both cases we can see that when the input is simulated
by its termination analyzer this input cannot possibly terminate
normally proves that these two cases have isomorphic essence.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/7/2024 8:19 PM, immibis wrote:
> On 1/8/24 03:12, olcott wrote:
>> On 1/6/2024 9:33 PM, immibis wrote:
>>> On 1/7/24 04:20, olcott wrote:
>>>> On 1/6/2024 7:23 PM, immibis wrote:
>>>>> On 1/7/24 01:24, olcott wrote:
>>>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>>>> On 1/6/24 21:48, olcott wrote:
>>>>>>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>>>>>>> On 1/6/24 19:32, olcott wrote:
>>>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>>>
>>>>>>>>>>>> Whether or not a question is incorrect is always determined
>>>>>>>>>>>> by ...
>>>>>>>>>>>
>>>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>>>>>> mistake.
>>>>>>>>>>
>>>>>>>>> What mistake?
>>>>>>>>
>>>>>>>> Undecidability is anchored in the fact that incorrect questions
>>>>>>>> have no correct answer. The correct approach is to reject these
>>>>>>>> incorrect questions and it is a mistake to not reject them.
>>>>>>>>
>>>>>>>
>>>>>>> Undecidability is actually the nonexistence of a Turing machine
>>>>>>> which always gives the correct answer to a certain question. That
>>>>>>> is the definition.
>>>>>>
>>>>>> When we screen out every question that cannot possibly have a
>>>>>> correct true/false return value as invalid input then undecidability
>>>>>> cannot possibly exist because these are the undecidable inputs.
>>>>>>
>>>>> The halting problem has a correct true/false return value
>>>>
>>>> *Not when the full context of the question is considered*
>>>> "What correct Boolean value does H return when D is defined
>>>>   to do the opposite of whatever value that H returns?"
>>>>
>>> That isn't the halting problem, that's a different problem.
>>
>> *That <is> the full context of the question that H is asked*
>
> We are talking about the halting problem, not H. The halting problem is:
> Does D(D) halt?
>
>> this IS NOT the same question that H is asked
>
> Well how do we ask H this question?
>
> If we can't ask H whether D(D) halts, then H doesn't solve the halting
> problem.

On 1/6/2024 1:54 PM, immibis wrote:
> "Does the barber shave himself?" has a correct answer: it's one of
> "yes" or "no".
>
> "Does a barber who shaves every man who does not shave himself shave
> himself?" has no correct answer.

*That was a brilliant insight*

*By that EXACT same reasoning*

Does the direct execution of D(D) halt? Has a correct answer: it is yes.

What correct Boolean value does H return when D is defined to do the
opposite of whatever value that H returns?" has no correct answer.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/7/24 10:18 PM, olcott wrote:
> On 1/6/2024 9:43 PM, immibis wrote:
>> On 1/7/24 04:00, olcott wrote:
>>> On 1/6/2024 7:21 PM, immibis wrote:
>>>> On 1/7/24 01:28, olcott wrote:
>>>>> On 1/6/2024 3:00 PM, immibis wrote:
>>>>>> On 1/6/24 21:32, olcott wrote:
>>>>>>>
>>>>>>> In that case embedded_H can recognize the its input is a copy of its
>>>>>>> own TM specification.
>>>>>>>
>>>>>>
>>>>>> The specification of embedded_H can be modified without altering
>>>>>> the semantics. It will give the same answer as before, but it will
>>>>>> no longer be a copy.
>>>>>
>>>>> There are several ways to recognize pathological self-reference.
>>>>> I presenting the simple one that has fully operational code.
>>>>> I could not encode the other one because the x86 emulator could
>>>>> not handle it. None-the-less it still does work on the Peter Linz
>>>>> Proof.
>>>>>
>>>> Your simple one isn't isomorphic to a Turing machine.
>>>
>>> It demonstrates a basic principle COMPLETELY without the gaps
>>> in reasoning that must be included in every HP proof.
>>>
>>> I show how the exact same basic principle <is> applied to the
>>> Peter Linz proof.
>>>
>>
>> Your simple on isn't isomorphic to a Turing machine, because of the
>> self-call-detection which has no analogue in a Turing machine.
>
> That in both cases we can see that when the input is simulated
> by its termination analyzer this input cannot possibly terminate
> normally proves that these two cases have isomorphic essence.
>

But, since the termination analyzer you are claiming gets the right
answer aborts its simulation, in never actually "Correctly SImulated"
its input.

And the simulation analyizer that didn't abort is simulation, isn't the
analyzer that the specific D we are considering was built from.

So, you are caught is a LIE, proving you just don't understand what a
progrtam is.

It seems, your logic is based on the "fact" that 1 is equal to 2, which
we all know is false.

If you want to talk about both cases, lets fully talk about both cases.

The H that answers, did so by aborting its simulation and returning the
non-halting answer. It thus never did a "Correct Simulation", and when
we do an actual correct simulation of this exact machine (which we can't
do with H) or just run it, we find that it halts.

For the other H, which your H assumes is its input (since you think 1 is
2) we see, as you point out, that this H never reaches a final state, so
this H just fails to be a decider. since the behavior of D is a function
of the H it is built on, this IS a different D, so this behavior doesn't
apply to the first case.

All you have done is prove you LIE that you H's answer is actually based
on a correct simultion done by H, since the H that answers didn't do a
correct simulation, and the correct simulaiton of the second machine,
was of a different input.

You are just proving your stupidity.

On 1/7/24 10:16 PM, olcott wrote:
> On 1/6/2024 9:42 PM, immibis wrote:
>> On 1/7/24 03:57, olcott wrote:
>>> On 1/6/2024 7:21 PM, immibis wrote:
>>>> On 1/7/24 01:45, olcott wrote:
>>>>  >
>>>>> The ZFC axiom schema of specification forbids the
>>>>> creation of self-referential sets.
>>>>>
>>>>
>>>> Yes, but you missed the forest for the trees. ZFC says that
>>>> self-referential sets do not exist.
>>>
>>> Not quite. ZFC prevents self-referential sets from coming into
>>> existence. Pathological self-reference (Olcott 2004) must be
>>> rejected as erroneous.
>>>
>>
>> Is it more obvious to say: Self-referential sets contradict the ZFC
>> axioms, so they can't exist in any world where the ZFC axioms are true.
>
> It is more important to say that every undecidable problem that is
> undecidable because of pathological self-reference is simply incorrect.

Which first means you have to actually show the "self-reference" in the
ACTUAL PROBLEM, and not just one in you lying Strawman.

Since Turing Machine can't have "references", you concept is just proved
incorrect.

>
>>>> Olcott machines say that self-referential programs do not exist. I
>>>> don't understand what you think is a self-referential program, but
>>>> we know how to construct one by following the steps of Linz (you
>>>> said this makes a self-referential program). So what happens when we
>>>> follow the steps? Is the result not a program?
>>>
>>
>> And I suppose self-referential programs contradict the Olcott axioms,
>> so they can't exist in any world where the Olcott axioms are true. The
>> problem is that I just wrote one, so obviously the Olcott axioms
>> aren't true in this world I live in.
>

On 1/7/24 10:21 PM, olcott wrote:
> On 1/7/2024 8:19 PM, immibis wrote:
>> On 1/8/24 03:12, olcott wrote:
>>> On 1/6/2024 9:33 PM, immibis wrote:
>>>> On 1/7/24 04:20, olcott wrote:
>>>>> On 1/6/2024 7:23 PM, immibis wrote:
>>>>>> On 1/7/24 01:24, olcott wrote:
>>>>>>> On 1/6/2024 2:58 PM, immibis wrote:
>>>>>>>> On 1/6/24 21:48, olcott wrote:
>>>>>>>>> On 1/6/2024 2:11 PM, immibis wrote:
>>>>>>>>>> On 1/6/24 19:32, olcott wrote:
>>>>>>>>>>> On 1/6/2024 10:24 AM, immibis wrote:
>>>>>>>>>>>> On 1/6/24 17:07, olcott wrote:
>>>>>>>>>>>>>
>>>>>>>>>>>>> Whether or not a question is incorrect is always determined
>>>>>>>>>>>>> by ...
>>>>>>>>>>>>
>>>>>>>>>>>> What is the purpose of talking about incorrect questions?
>>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> Proving that the whole notion of undecidability is anchored in a
>>>>>>>>>>> mistake.
>>>>>>>>>>>
>>>>>>>>>> What mistake?
>>>>>>>>>
>>>>>>>>> Undecidability is anchored in the fact that incorrect questions
>>>>>>>>> have no correct answer. The correct approach is to reject these
>>>>>>>>> incorrect questions and it is a mistake to not reject them.
>>>>>>>>>
>>>>>>>>
>>>>>>>> Undecidability is actually the nonexistence of a Turing machine
>>>>>>>> which always gives the correct answer to a certain question.
>>>>>>>> That is the definition.
>>>>>>>
>>>>>>> When we screen out every question that cannot possibly have a
>>>>>>> correct true/false return value as invalid input then undecidability
>>>>>>> cannot possibly exist because these are the undecidable inputs.
>>>>>>>
>>>>>> The halting problem has a correct true/false return value
>>>>>
>>>>> *Not when the full context of the question is considered*
>>>>> "What correct Boolean value does H return when D is defined
>>>>>   to do the opposite of whatever value that H returns?"
>>>>>
>>>> That isn't the halting problem, that's a different problem.
>>>
>>> *That <is> the full context of the question that H is asked*
>>
>> We are talking about the halting problem, not H. The halting problem
>> is: Does D(D) halt?
>>
>>> this IS NOT the same question that H is asked
>>
>> Well how do we ask H this question?
>>
>> If we can't ask H whether D(D) halts, then H doesn't solve the halting
>> problem.
>
> On 1/6/2024 1:54 PM, immibis wrote:
> > "Does the barber shave himself?" has a correct answer: it's one of
> > "yes" or "no".
> >
> > "Does a barber who shaves every man who does not shave himself shave
> > himself?" has no correct answer.
>
> *That was a brilliant insight*
>
> *By that EXACT same reasoning*
>
> Does the direct execution of D(D) halt? Has a correct answer: it is yes.
>
> What correct Boolean value does H return when D is defined to do the
> opposite of whatever value that H returns?" has no correct answer.
>
>

No, it just shows that like the Barber problem, the error is in defining
the decider to have impossilbe charactersitics, namely to be able toe
get the answer correct were the QUESTION adds a self-reference that
isn't in the original.

Does D(D) Halt, has no "self-reference" in it (or even a reference to H).

Adding the what can H return to be correct, adds the error in the
assumption that H can return something to be correct.

THe original question still has a correct answer for any H that actually
exists.

The modified question doesn't have an H that meets it that can exist, so
is invalid.

On 1/7/2024 9:58 PM, Richard Damon wrote:
> On 1/7/24 10:18 PM, olcott wrote:
>> On 1/6/2024 9:43 PM, immibis wrote:
>>> On 1/7/24 04:00, olcott wrote:
>>>> On 1/6/2024 7:21 PM, immibis wrote:
>>>>> On 1/7/24 01:28, olcott wrote:
>>>>>> On 1/6/2024 3:00 PM, immibis wrote:
>>>>>>> On 1/6/24 21:32, olcott wrote:
>>>>>>>>
>>>>>>>> In that case embedded_H can recognize the its input is a copy of
>>>>>>>> its
>>>>>>>> own TM specification.
>>>>>>>>
>>>>>>>
>>>>>>> The specification of embedded_H can be modified without altering
>>>>>>> the semantics. It will give the same answer as before, but it
>>>>>>> will no longer be a copy.
>>>>>>
>>>>>> There are several ways to recognize pathological self-reference.
>>>>>> I presenting the simple one that has fully operational code.
>>>>>> I could not encode the other one because the x86 emulator could
>>>>>> not handle it. None-the-less it still does work on the Peter Linz
>>>>>> Proof.
>>>>>>
>>>>> Your simple one isn't isomorphic to a Turing machine.
>>>>
>>>> It demonstrates a basic principle COMPLETELY without the gaps
>>>> in reasoning that must be included in every HP proof.
>>>>
>>>> I show how the exact same basic principle <is> applied to the
>>>> Peter Linz proof.
>>>>
>>>
>>> Your simple on isn't isomorphic to a Turing machine, because of the
>>> self-call-detection which has no analogue in a Turing machine.
>>
>> That in both cases we can see that when the input is simulated
>> by its termination analyzer this input cannot possibly terminate
>> normally proves that these two cases have isomorphic essence.
>>
>
> But, since the termination analyzer you are claiming

No simulating termination analyzer that can possibly exist
such that the input to H(D,D) terminates normally or
the input to embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ terminates normally.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

On 1/8/24 05:14, olcott wrote:
> On 1/7/2024 9:58 PM, Richard Damon wrote:
>> On 1/7/24 10:18 PM, olcott wrote:
>>> On 1/6/2024 9:43 PM, immibis wrote:
>>>> On 1/7/24 04:00, olcott wrote:
>>>>> On 1/6/2024 7:21 PM, immibis wrote:
>>>>>> On 1/7/24 01:28, olcott wrote:
>>>>>>> On 1/6/2024 3:00 PM, immibis wrote:
>>>>>>>> On 1/6/24 21:32, olcott wrote:
>>>>>>>>>
>>>>>>>>> In that case embedded_H can recognize the its input is a copy
>>>>>>>>> of its
>>>>>>>>> own TM specification.
>>>>>>>>>
>>>>>>>>
>>>>>>>> The specification of embedded_H can be modified without altering
>>>>>>>> the semantics. It will give the same answer as before, but it
>>>>>>>> will no longer be a copy.
>>>>>>>
>>>>>>> There are several ways to recognize pathological self-reference.
>>>>>>> I presenting the simple one that has fully operational code.
>>>>>>> I could not encode the other one because the x86 emulator could
>>>>>>> not handle it. None-the-less it still does work on the Peter Linz
>>>>>>> Proof.
>>>>>>>
>>>>>> Your simple one isn't isomorphic to a Turing machine.
>>>>>
>>>>> It demonstrates a basic principle COMPLETELY without the gaps
>>>>> in reasoning that must be included in every HP proof.
>>>>>
>>>>> I show how the exact same basic principle <is> applied to the
>>>>> Peter Linz proof.
>>>>>
>>>>
>>>> Your simple on isn't isomorphic to a Turing machine, because of the
>>>> self-call-detection which has no analogue in a Turing machine.
>>>
>>> That in both cases we can see that when the input is simulated
>>> by its termination analyzer this input cannot possibly terminate
>>> normally proves that these two cases have isomorphic essence.
>>>
>>
>> But, since the termination analyzer you are claiming
>
> No simulating termination analyzer that can possibly exist
> such that the input to H(D,D) terminates normally or
> the input to embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ terminates normally.
>

No simulating termination analyzer can possibly exist.

On 1/8/24 03:45, olcott wrote:
> On 1/6/2024 1:54 PM, immibis wrote:
>> "Does the barber shave himself?" has a correct answer: it's one of
>> "yes" or "no".
>>
>> "Does a barber who shaves every man who does not shave himself shave
>> himself?" has no correct answer.
>
> *That was a brilliant insight*
>
> *By that EXACT same reasoning*
>
> Does the direct execution of D(D) halt? Has a correct answer: it is yes.
>
> What correct Boolean value does H return when D is defined to do the
> opposite of whatever value that H returns?" has no correct answer.
>

That's right. There is no answer to "What correct Boolean value does H
return?" because H returns an incorrect Boolean value.

It is like asking "which kind of fish does the barber shave?"

On 1/7/2024 11:06 PM, immibis wrote:
> On 1/8/24 05:14, olcott wrote:
>> On 1/7/2024 9:58 PM, Richard Damon wrote:
>>> On 1/7/24 10:18 PM, olcott wrote:
>>>> On 1/6/2024 9:43 PM, immibis wrote:
>>>>> On 1/7/24 04:00, olcott wrote:
>>>>>> On 1/6/2024 7:21 PM, immibis wrote:
>>>>>>> On 1/7/24 01:28, olcott wrote:
>>>>>>>> On 1/6/2024 3:00 PM, immibis wrote:
>>>>>>>>> On 1/6/24 21:32, olcott wrote:
>>>>>>>>>>
>>>>>>>>>> In that case embedded_H can recognize the its input is a copy
>>>>>>>>>> of its
>>>>>>>>>> own TM specification.
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>> The specification of embedded_H can be modified without
>>>>>>>>> altering the semantics. It will give the same answer as before,
>>>>>>>>> but it will no longer be a copy.
>>>>>>>>
>>>>>>>> There are several ways to recognize pathological self-reference.
>>>>>>>> I presenting the simple one that has fully operational code.
>>>>>>>> I could not encode the other one because the x86 emulator could
>>>>>>>> not handle it. None-the-less it still does work on the Peter
>>>>>>>> Linz Proof.
>>>>>>>>
>>>>>>> Your simple one isn't isomorphic to a Turing machine.
>>>>>>
>>>>>> It demonstrates a basic principle COMPLETELY without the gaps
>>>>>> in reasoning that must be included in every HP proof.
>>>>>>
>>>>>> I show how the exact same basic principle <is> applied to the
>>>>>> Peter Linz proof.
>>>>>>
>>>>>
>>>>> Your simple on isn't isomorphic to a Turing machine, because of the
>>>>> self-call-detection which has no analogue in a Turing machine.
>>>>
>>>> That in both cases we can see that when the input is simulated
>>>> by its termination analyzer this input cannot possibly terminate
>>>> normally proves that these two cases have isomorphic essence.
>>>>
>>>
>>> But, since the termination analyzer you are claiming
>>
>> No simulating termination analyzer that can possibly exist
>> such that the input to H(D,D) terminates normally or
>> the input to embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ terminates normally.
>>
>
> No simulating termination analyzer can possibly exist.
https://github.com/plolcott/x86utm

Is fully operational and correctly determines non-halting
for many combinations of simple infinite loops, simple
infinite recursion and simple recursive simulation.

--
Copyright 2023 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer