• Re: An idea for a simulating halt decider

    From olcott@21:1/5 to Ben Bacarisse on Mon Jul 4 19:01:58 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/4/2022 6:17 PM, Ben Bacarisse wrote:
    Mr Flibble <[email protected]> writes:

    On Mon, 04 Jul 2022 02:27:58 +0100
    Ben Bacarisse <[email protected]> wrote:

    Mr Flibble <[email protected]> writes:

    I hereby assert my copyright for the forking simulating halt
    decider. :)

    I note the smiley, but you can't copyright an idea. Copyright
    pertains to forms of expression, even when the idea is not the
    author's. I will have copyright on any sufficiently novel exposition
    I might publish about your ideas!

    I published my idea in this forum and that exposition is what I am
    copyrighting.

    OK. I'm just pointing out that there's little point in doing that. The exposition is probably not what you care about.


    I publish to this forum primarily to establish original authorship of my
    ideas. Even though everyone here had only reviewed my work for the
    purpose of finding fault and or rebuttal these reviews were crucially
    required for me to achieve all of the progress that I have made.

    Now that my work is entirely proven to be correct entirely on the basis
    of established facts:

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    Competent reviewers are no longer willing to review it.

    *Halting problem proofs refuted on the basis of software engineering*

    https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering

    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Richard Damon@21:1/5 to olcott on Tue Jul 5 07:32:37 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/4/22 8:01 PM, olcott wrote:
    On 7/4/2022 6:17 PM, Ben Bacarisse wrote:
    Mr Flibble <[email protected]> writes:

    On Mon, 04 Jul 2022 02:27:58 +0100
    Ben Bacarisse <[email protected]> wrote:

    Mr Flibble <[email protected]> writes:

    I hereby assert my copyright for the forking simulating halt
    decider. :)

    I note the smiley, but you can't copyright an idea.  Copyright
    pertains to forms of expression, even when the idea is not the
    author's.  I will have copyright on any sufficiently novel exposition >>>> I might publish about your ideas!
    I published my idea in this forum and that exposition is what I am
    copyrighting.

    OK.  I'm just pointing out that there's little point in doing that.  The >> exposition is probably not what you care about.


    I publish to this forum primarily to establish original authorship of my ideas. Even though everyone here had only reviewed my work for the
    purpose of finding fault and or rebuttal these reviews were crucially required for me to achieve all of the progress that I have made.

    Now that my work is entirely proven to be correct entirely on the basis
    of established facts:

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    Competent reviewers are no longer willing to review it.

    *Halting problem proofs refuted on the basis of software engineering*

    https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering



    Except it has been pointed out that it doesn't, because H doesn't
    actually do a correct and complete x86 emulation, so you can't prove
    anything based on the thing it didn't do.

    That would be like saying since Trump won the 2020 election, he should
    have appointed the latest Supreme Court replacement.

    If you think you actually HAVE proven it, why not go away and submit the
    paper?

    The answer is that you know your "proof" isn't true, and need to find
    better words to hide that fact.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Andy Walker on Wed Jul 6 08:32:05 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much
    advanced and
    promising than the oral-based halting decider (POOH). Chance might be
    good
    refuting the HP.

        If by "refuting the HP" you mean "constructing a halt decider", then you have as much chance as refuting that 2+2 == 4 [in all cases, with the usual meanings of those words].  All the obfuscation of the last couple of decades does absolutely nothing to indicate any actual error in any of the several known proofs that no general halt decider can exist.  If you, or
    PO,
    ever did manage to produce an actual purported "H", then we already know
    how
    to construct an actual counterexample that refutes your, or his, claim. That's all anyone really needs to know.  We can sit back and wait however long it takes for an actual claimed "H" [whether in C or x86 code or as a
    TM] to appear, and then it is a matter of moments to produce a program and input that "H" fails with.

        If by "refuting the HP" you mean something else, then you need to explain further, as "refuting" in English applies to claims rather than
    to problems.


    I dare you to try to refute this.

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    void P(u32 x)
    {
    if (H(x, x))
    HERE: goto HERE;
    return;
    }

    int main()
    {
    Output("Input_Halts = ", H((u32)P, (u32)P));
    }

    For any program H that might determine if programs halt, a
    "pathological"
    program P, called with some input, can pass its own source and its
    input to
    H and then specifically do the opposite of what H predicts P will
    do. No H
    can exist that handles this case. https://en.wikipedia.org/wiki/Halting_problem

    H and P implement the above specified pathological relationship to each
    other:


    *Halting problem proofs refuted on the basis of software engineering* https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering


    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Dennis Bush on Wed Jul 6 09:00:56 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 8:37 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 9:32:13 AM UTC-4, olcott wrote:
    On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much
    advanced and
    promising than the oral-based halting decider (POOH). Chance might be
    good
    refuting the HP.

    If by "refuting the HP" you mean "constructing a halt decider", then >>> you have as much chance as refuting that 2+2 == 4 [in all cases, with the >>> usual meanings of those words]. All the obfuscation of the last couple of >>> decades does absolutely nothing to indicate any actual error in any of the >>> several known proofs that no general halt decider can exist. If you, or >>> PO,
    ever did manage to produce an actual purported "H", then we already know >>> how
    to construct an actual counterexample that refutes your, or his, claim.
    That's all anyone really needs to know. We can sit back and wait however >>> long it takes for an actual claimed "H" [whether in C or x86 code or as a >>> TM] to appear, and then it is a matter of moments to produce a program and >>> input that "H" fails with.

    If by "refuting the HP" you mean something else, then you need to
    explain further, as "refuting" in English applies to claims rather than
    to problems.

    I dare you to try to refute this.
    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    H has a fixed algorithm which we'll call Ha. And since P calls Pa, we'll refer to it as Pa. So Ha is doing the deciding and Pa is what is being decided on.

    Because the fixed algorithm of Ha aborts, it does not do a correct and complete emulation. Therefore "[Ha's] correct and complete x86 emulation of its input" does not exist, making the above statement nonsense.

    This has been told to you several times before and you have not attempted to explain why it is wrong. Therefore anyone that reads this is forced to conclude that you are unable to and therefore agree that your statement is nonsense.

    In other words Dennis and Richard are saying that no halt decider can
    ever correctly determine (in a finite number of steps) that
    Infinite_Loop() never halts.

    *H0 correctly determines that Infinite_Loop() never halts*

    void Infinite_Loop()
    {
    HERE: goto HERE;
    }

    int main()
    {
    Output("Input_Halts = ", H0((u32)Infinite_Loop));
    }

    _Infinite_Loop()
    [00001102](01) 55 push ebp
    [00001103](02) 8bec mov ebp,esp
    [00001105](02) ebfe jmp 00001105
    [00001107](01) 5d pop ebp
    [00001108](01) c3 ret
    Size in bytes:(0007) [00001108]

    _main()
    [00001192](01) 55 push ebp
    [00001193](02) 8bec mov ebp,esp
    [00001195](05) 6802110000 push 00001102
    [0000119a](05) e8d3fbffff call 00000d72
    [0000119f](03) 83c404 add esp,+04
    [000011a2](01) 50 push eax
    [000011a3](05) 68a3040000 push 000004a3
    [000011a8](05) e845f3ffff call 000004f2
    [000011ad](03) 83c408 add esp,+08
    [000011b0](02) 33c0 xor eax,eax
    [000011b2](01) 5d pop ebp
    [000011b3](01) c3 ret
    Size in bytes:(0034) [000011b3]

    machine stack stack machine assembly
    address address data code language
    ======== ======== ======== ========= ============= [00001192][00101ef8][00000000] 55 push ebp [00001193][00101ef8][00000000] 8bec mov ebp,esp [00001195][00101ef4][00001102] 6802110000 push 00001102 [0000119a][00101ef0][0000119f] e8d3fbffff call 00000d72

    H0: Begin Simulation Execution Trace Stored at:211fac [00001102][00211f9c][00211fa0] 55 push ebp [00001103][00211f9c][00211fa0] 8bec mov ebp,esp [00001105][00211f9c][00211fa0] ebfe jmp 00001105 [00001105][00211f9c][00211fa0] ebfe jmp 00001105
    H0: Infinite Loop Detected Simulation Stopped

    if (current->Simplified_Opcode == JMP) // JMP
    if (current->Decode_Target <= current->Address) // upward
    if (traced->Address == current->Decode_Target) // to this address
    if (Conditional_Branch_Count == 0) // no escape
    return 1;

    [0000119f][00101ef8][00000000] 83c404 add esp,+04 [000011a2][00101ef4][00000000] 50 push eax [000011a3][00101ef0][000004a3] 68a3040000 push 000004a3 [000011a8][00101ef0][000004a3] e845f3ffff call 000004f2
    Input_Halts = 0
    [000011ad][00101ef8][00000000] 83c408 add esp,+08 [000011b0][00101ef8][00000000] 33c0 xor eax,eax [000011b2][00101efc][00100000] 5d pop ebp [000011b3][00101f00][00000004] c3 ret
    Number of Instructions Executed(554) == 8 Pages



    *Halting problem proofs refuted on the basis of software engineering* https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering


    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Dennis Bush on Wed Jul 6 09:28:02 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 9:06 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:01:06 AM UTC-4, olcott wrote:
    On 7/6/2022 8:37 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 9:32:13 AM UTC-4, olcott wrote:
    On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much
    advanced and
    promising than the oral-based halting decider (POOH). Chance might be >>>>>> good
    refuting the HP.

    If by "refuting the HP" you mean "constructing a halt decider", then >>>>> you have as much chance as refuting that 2+2 == 4 [in all cases, with the >>>>> usual meanings of those words]. All the obfuscation of the last couple of >>>>> decades does absolutely nothing to indicate any actual error in any of the
    several known proofs that no general halt decider can exist. If you, or >>>>> PO,
    ever did manage to produce an actual purported "H", then we already know >>>>> how
    to construct an actual counterexample that refutes your, or his, claim. >>>>> That's all anyone really needs to know. We can sit back and wait however >>>>> long it takes for an actual claimed "H" [whether in C or x86 code or as a >>>>> TM] to appear, and then it is a matter of moments to produce a program and
    input that "H" fails with.

    If by "refuting the HP" you mean something else, then you need to
    explain further, as "refuting" in English applies to claims rather than >>>>> to problems.

    I dare you to try to refute this.
    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would >>>> never reach the "ret" instruction (final state) of this input.

    H has a fixed algorithm which we'll call Ha. And since P calls Pa, we'll refer to it as Pa. So Ha is doing the deciding and Pa is what is being decided on.

    Because the fixed algorithm of Ha aborts, it does not do a correct and complete emulation. Therefore "[Ha's] correct and complete x86 emulation of its input" does not exist, making the above statement nonsense.

    This has been told to you several times before and you have not attempted to explain why it is wrong. Therefore anyone that reads this is forced to conclude that you are unable to and therefore agree that your statement is nonsense.
    In other words Dennis and Richard are saying that no halt decider can
    ever correctly determine (in a finite number of steps) that
    Infinite_Loop() never halts.

    Infinite_Loop has previously shown to be irrelevant.

    Try again.


    Your claim is that it is impossible for a halt decider to correctly
    predict that its input would never halt unless this input is simulated
    forever and it never halts. I proved that you are wrong about this.

    H0(Infinite_Loop);
    H(Infinite_Recursion, 0x777);
    H(P,P);

    All correctly predict (in a finite number of steps) that their complete
    and correct x86 emulation of their input would never reach the "ret" instruction of this input.

    The above three examples are provided as Example_01, Example_02, and
    Example_03 in my linked paper below.

    *Halting problem proofs refuted on the basis of software engineering* https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering


    --

    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Dennis Bush on Wed Jul 6 10:05:04 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 9:44 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:28:10 AM UTC-4, olcott wrote:
    On 7/6/2022 9:06 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:01:06 AM UTC-4, olcott wrote:
    On 7/6/2022 8:37 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 9:32:13 AM UTC-4, olcott wrote:
    On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much >>>>>>>> advanced and
    promising than the oral-based halting decider (POOH). Chance might be >>>>>>>> good
    refuting the HP.

    If by "refuting the HP" you mean "constructing a halt decider", then >>>>>>> you have as much chance as refuting that 2+2 == 4 [in all cases, with the
    usual meanings of those words]. All the obfuscation of the last couple of
    decades does absolutely nothing to indicate any actual error in any of the
    several known proofs that no general halt decider can exist. If you, or >>>>>>> PO,
    ever did manage to produce an actual purported "H", then we already know
    how
    to construct an actual counterexample that refutes your, or his, claim. >>>>>>> That's all anyone really needs to know. We can sit back and wait however
    long it takes for an actual claimed "H" [whether in C or x86 code or as a
    TM] to appear, and then it is a matter of moments to produce a program and
    input that "H" fails with.

    If by "refuting the HP" you mean something else, then you need to >>>>>>> explain further, as "refuting" in English applies to claims rather than >>>>>>> to problems.

    I dare you to try to refute this.
    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would >>>>>> never reach the "ret" instruction (final state) of this input.

    H has a fixed algorithm which we'll call Ha. And since P calls Pa, we'll refer to it as Pa. So Ha is doing the deciding and Pa is what is being decided on.

    Because the fixed algorithm of Ha aborts, it does not do a correct and complete emulation. Therefore "[Ha's] correct and complete x86 emulation of its input" does not exist, making the above statement nonsense.

    This has been told to you several times before and you have not attempted to explain why it is wrong. Therefore anyone that reads this is forced to conclude that you are unable to and therefore agree that your statement is nonsense.
    In other words Dennis and Richard are saying that no halt decider can
    ever correctly determine (in a finite number of steps) that
    Infinite_Loop() never halts.

    Infinite_Loop has previously shown to be irrelevant.

    Try again.

    Your claim is that it is impossible for a halt decider to correctly
    predict that its input would never halt unless this input is simulated
    forever and it never halts. I proved that you are wrong about this.

    That's not what I said. I said that it's impossible for Ha(Pa,Pa) to do a correct and complete emulation of its input because it aborts, therefore its nonsense to predict what Ha(Pa,Pa)'s correct and complete emulation of its input will do.

    The halt deciders all correctly predict (in a finite number of steps)
    what their correct and complete emulation of their input would do on the
    basis that these inputs demonstrate non-halting behavior patterns that
    the halt deciders recognize.

    If your nonsense reasoning applies to H(P,P) then when this same
    nonsense reasoning is applied to H0(Infinite_Loop) and
    H(Infinite_Recursion, 0x777) it is shown to be nonsense.

    All three of these examples are provided in my paper:
    Example_01 H0(Infinite_Loop);
    Example_02 H(Infinite_Recursion, 0x777);
    Example_03 H(P,P);

    *Halting problem proofs refuted on the basis of software engineering* https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering

    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Dennis Bush on Wed Jul 6 10:54:04 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 10:16 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 11:05:12 AM UTC-4, olcott wrote:
    On 7/6/2022 9:44 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:28:10 AM UTC-4, olcott wrote:
    On 7/6/2022 9:06 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:01:06 AM UTC-4, olcott wrote:
    On 7/6/2022 8:37 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 9:32:13 AM UTC-4, olcott wrote:
    On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much >>>>>>>>>> advanced and
    promising than the oral-based halting decider (POOH). Chance might be
    good
    refuting the HP.

    If by "refuting the HP" you mean "constructing a halt decider", then >>>>>>>>> you have as much chance as refuting that 2+2 == 4 [in all cases, with the
    usual meanings of those words]. All the obfuscation of the last couple of
    decades does absolutely nothing to indicate any actual error in any of the
    several known proofs that no general halt decider can exist. If you, or
    PO,
    ever did manage to produce an actual purported "H", then we already know
    how
    to construct an actual counterexample that refutes your, or his, claim.
    That's all anyone really needs to know. We can sit back and wait however
    long it takes for an actual claimed "H" [whether in C or x86 code or as a
    TM] to appear, and then it is a matter of moments to produce a program and
    input that "H" fails with.

    If by "refuting the HP" you mean something else, then you need to >>>>>>>>> explain further, as "refuting" in English applies to claims rather than
    to problems.

    I dare you to try to refute this.
    From a purely software engineering perspective (anchored in the >>>>>>>> semantics of the x86 language) it is proven that H(P,P) correctly >>>>>>>> predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    H has a fixed algorithm which we'll call Ha. And since P calls Pa, we'll refer to it as Pa. So Ha is doing the deciding and Pa is what is being decided on.

    Because the fixed algorithm of Ha aborts, it does not do a correct and complete emulation. Therefore "[Ha's] correct and complete x86 emulation of its input" does not exist, making the above statement nonsense.

    This has been told to you several times before and you have not attempted to explain why it is wrong. Therefore anyone that reads this is forced to conclude that you are unable to and therefore agree that your statement is nonsense.
    In other words Dennis and Richard are saying that no halt decider can >>>>>> ever correctly determine (in a finite number of steps) that
    Infinite_Loop() never halts.

    Infinite_Loop has previously shown to be irrelevant.

    Try again.

    Your claim is that it is impossible for a halt decider to correctly
    predict that its input would never halt unless this input is simulated >>>> forever and it never halts. I proved that you are wrong about this.

    That's not what I said. I said that it's impossible for Ha(Pa,Pa) to do a correct and complete emulation of its input because it aborts, therefore its nonsense to predict what Ha(Pa,Pa)'s correct and complete emulation of its input will do.
    The halt deciders all correctly predict (in a finite number of steps)
    what their correct and complete emulation of their input would do on the
    basis that these inputs demonstrate non-halting behavior patterns that
    the halt deciders recognize.

    But in the case of Ha(Pa,Pa) it *doesn't* do a correct and complete emulation of its input, so it makes no sense to say what Ha's correct and complete emulation would be.


    If your nonsense reasoning applies to H(P,P) then when this same
    nonsense reasoning is applied to H0(Infinite_Loop) and
    H(Infinite_Recursion, 0x777) it is shown to be nonsense.

    Yes, it applies to those as well as none of them do a complete and correct simulation of their inputs. They happen to get the right answer because an *actual* correct and complete simulation of their inputs, i.e. UTM(Infinite_Loop) and UTM(Infinite_
    Recursion, 0x777) do not halt.

    The correct and complete simulation of the input to Ha(Pa,Pa),

    A simulating halt decider correctly predicts whether or not its correct
    and complete simulation of its input would ever reach the final state of
    this input.

    Many times its correct and complete simulation of its input would match
    the behavior of another different simulator simulating this same input.

    For any program H that might determine if programs halt, a
    "pathological"
    program P, called with some input, can pass its own source and its
    input to
    H and then specifically do the opposite of what H predicts P will
    do. No H
    can exist that handles this case. https://en.wikipedia.org/wiki/Halting_problem

    This is not the case when H and P have the above halting theorem
    pathological relationship to each other. In that case it is the actual
    behavior of the actual input that must be assessed.

    A halt decider must compute the mapping from its inputs to an accept or
    reject state on the basis of the actual behavior that is actually
    specified by these inputs.

    P(P) reaches its "ret" instruction.

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    That you insist that infinite behavior can only be recognized by
    infinite simulation is a bald faced lie and you know it. I have proven otherwise by H0(Infinite_Loop) and H(Infinite_Recursion, 0x777).


    i.e. UTM(Pa,Pa), *does* halt, so Ha(Pa,Pa)==0 is wrong


    All three of these examples are provided in my paper:
    Example_01 H0(Infinite_Loop);
    Example_02 H(Infinite_Recursion, 0x777);
    Example_03 H(P,P);
    *Halting problem proofs refuted on the basis of software engineering*
    https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering

    --
    Copyright 2022 Pete Olcott

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


    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Dennis Bush on Wed Jul 6 11:28:40 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 11:19 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 12:15:52 PM UTC-4, olcott wrote:
    On 7/6/2022 11:09 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 11:54:12 AM UTC-4, olcott wrote:
    On 7/6/2022 10:16 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 11:05:12 AM UTC-4, olcott wrote:
    On 7/6/2022 9:44 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:28:10 AM UTC-4, olcott wrote:
    On 7/6/2022 9:06 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:01:06 AM UTC-4, olcott wrote: >>>>>>>>>> On 7/6/2022 8:37 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 9:32:13 AM UTC-4, olcott wrote: >>>>>>>>>>>> On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much >>>>>>>>>>>>>> advanced and
    promising than the oral-based halting decider (POOH). Chance might be
    good
    refuting the HP.

    If by "refuting the HP" you mean "constructing a halt decider", then
    you have as much chance as refuting that 2+2 == 4 [in all cases, with the
    usual meanings of those words]. All the obfuscation of the last couple of
    decades does absolutely nothing to indicate any actual error in any of the
    several known proofs that no general halt decider can exist. If you, or
    PO,
    ever did manage to produce an actual purported "H", then we already know
    how
    to construct an actual counterexample that refutes your, or his, claim.
    That's all anyone really needs to know. We can sit back and wait however
    long it takes for an actual claimed "H" [whether in C or x86 code or as a
    TM] to appear, and then it is a matter of moments to produce a program and
    input that "H" fails with.

    If by "refuting the HP" you mean something else, then you need to >>>>>>>>>>>>> explain further, as "refuting" in English applies to claims rather than
    to problems.

    I dare you to try to refute this.
    From a purely software engineering perspective (anchored in the >>>>>>>>>>>> semantics of the x86 language) it is proven that H(P,P) correctly >>>>>>>>>>>> predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input. >>>>>>>>>>>
    H has a fixed algorithm which we'll call Ha. And since P calls Pa, we'll refer to it as Pa. So Ha is doing the deciding and Pa is what is being decided on.

    Because the fixed algorithm of Ha aborts, it does not do a correct and complete emulation. Therefore "[Ha's] correct and complete x86 emulation of its input" does not exist, making the above statement nonsense.

    This has been told to you several times before and you have not attempted to explain why it is wrong. Therefore anyone that reads this is forced to conclude that you are unable to and therefore agree that your statement is nonsense.
    In other words Dennis and Richard are saying that no halt decider can
    ever correctly determine (in a finite number of steps) that >>>>>>>>>> Infinite_Loop() never halts.

    Infinite_Loop has previously shown to be irrelevant.

    Try again.

    Your claim is that it is impossible for a halt decider to correctly >>>>>>>> predict that its input would never halt unless this input is simulated >>>>>>>> forever and it never halts. I proved that you are wrong about this. >>>>>>>
    That's not what I said. I said that it's impossible for Ha(Pa,Pa) to do a correct and complete emulation of its input because it aborts, therefore its nonsense to predict what Ha(Pa,Pa)'s correct and complete emulation of its input will do.
    The halt deciders all correctly predict (in a finite number of steps) >>>>>> what their correct and complete emulation of their input would do on the >>>>>> basis that these inputs demonstrate non-halting behavior patterns that >>>>>> the halt deciders recognize.

    But in the case of Ha(Pa,Pa) it *doesn't* do a correct and complete emulation of its input, so it makes no sense to say what Ha's correct and complete emulation would be.


    If your nonsense reasoning applies to H(P,P) then when this same
    nonsense reasoning is applied to H0(Infinite_Loop) and
    H(Infinite_Recursion, 0x777) it is shown to be nonsense.

    Yes, it applies to those as well as none of them do a complete and correct simulation of their inputs. They happen to get the right answer because an *actual* correct and complete simulation of their inputs, i.e. UTM(Infinite_Loop) and UTM(Infinite_
    Recursion, 0x777) do not halt.

    The correct and complete simulation of the input to Ha(Pa,Pa),
    A simulating halt decider correctly predicts whether or not its correct >>>> and complete simulation of its input would ever reach the final state of >>>> this input.

    So you again just repeat your claim without explaining why I am wrong.

    Which means you agree that your claim is nonsense.

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input

    Why do you continue to make this claim after you've agreed that it's nonsense?


    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P)

    correctly predicts (without doing a complete emulation)
    correctly predicts (without doing a complete emulation)
    correctly predicts (without doing a complete emulation)
    correctly predicts (without doing a complete emulation)

    that its correct and complete x86 emulation of its input would never
    reach the "ret" instruction (final state) of this input.

    It does this by correctly recognizing an infinite behavior pattern in
    its partial x86 emulation of its input.

    It does this by correctly recognizing an infinite behavior pattern in
    its partial x86 emulation of its input.

    It does this by correctly recognizing an infinite behavior pattern in
    its partial x86 emulation of its input.

    It does this by correctly recognizing an infinite behavior pattern in
    its partial x86 emulation of its input.


    Ha(Pa,Pa) does not perform a correct and complete simulation, therefore Ha(Pa,Pa)'s correct and complete simulation does not exist.


    would
    never reach the "ret" instruction (final state) of this input.
    A halt decider computes the mapping from its inputs to an accept or
    reject state based on the actual behavior of these actual inputs.

    H(P,P) does do that therefore H(P,P)==0 is necessarily correct. Every
    rebuttal of a necessarily correct statement is necessarily incorrect.

    Many times its correct and complete simulation of its input

    Simply doesn't exist if the decider aborts

    would match
    the behavior of another different simulator simulating this same input. >>>> For any program H that might determine if programs halt, a
    "pathological"
    program P, called with some input, can pass its own source and its
    input to
    H and then specifically do the opposite of what H predicts P will
    do. No H
    can exist that handles this case.
    https://en.wikipedia.org/wiki/Halting_problem
    This is not the case when H and P have the above halting theorem
    pathological relationship to each other. In that case it is the actual >>>> behavior of the actual input that must be assessed.

    A halt decider must compute the mapping from its inputs to an accept or >>>> reject state on the basis of the actual behavior that is actually
    specified by these inputs.

    Which by the specification put forth by the halting problem:

    H(x,y)==1 if and only if x(y) halts, and
    H(x,y)==0 if and only if x(y) does not halt

    Is the behavior of x(y)

    Remember, the halting problem is asking the question "Does an *algorithm* exist that can determine if *any* arbitrary algorithm will halt given a particular input", and *if* such an algorithm exits, it MUST implement the above specification.


    P(P) reaches its "ret" instruction.
    From a purely software engineering perspective (anchored in the
    semantics of the x86 language)

    Ha must implement the specification it is stipulated to, which is

    H(x,y)==1 if and only if x(y) halts, and
    H(x,y)==0 if and only if x(y) does not halt

    it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input

    Does not exist if Ha aborts



    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From olcott@21:1/5 to Dennis Bush on Wed Jul 6 11:15:44 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/2022 11:09 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 11:54:12 AM UTC-4, olcott wrote:
    On 7/6/2022 10:16 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 11:05:12 AM UTC-4, olcott wrote:
    On 7/6/2022 9:44 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:28:10 AM UTC-4, olcott wrote:
    On 7/6/2022 9:06 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 10:01:06 AM UTC-4, olcott wrote:
    On 7/6/2022 8:37 AM, Dennis Bush wrote:
    On Wednesday, July 6, 2022 at 9:32:13 AM UTC-4, olcott wrote: >>>>>>>>>> On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much >>>>>>>>>>>> advanced and
    promising than the oral-based halting decider (POOH). Chance might be
    good
    refuting the HP.

    If by "refuting the HP" you mean "constructing a halt decider", then
    you have as much chance as refuting that 2+2 == 4 [in all cases, with the
    usual meanings of those words]. All the obfuscation of the last couple of
    decades does absolutely nothing to indicate any actual error in any of the
    several known proofs that no general halt decider can exist. If you, or
    PO,
    ever did manage to produce an actual purported "H", then we already know
    how
    to construct an actual counterexample that refutes your, or his, claim.
    That's all anyone really needs to know. We can sit back and wait however
    long it takes for an actual claimed "H" [whether in C or x86 code or as a
    TM] to appear, and then it is a matter of moments to produce a program and
    input that "H" fails with.

    If by "refuting the HP" you mean something else, then you need to >>>>>>>>>>> explain further, as "refuting" in English applies to claims rather than
    to problems.

    I dare you to try to refute this.
    From a purely software engineering perspective (anchored in the >>>>>>>>>> semantics of the x86 language) it is proven that H(P,P) correctly >>>>>>>>>> predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input. >>>>>>>>>
    H has a fixed algorithm which we'll call Ha. And since P calls Pa, we'll refer to it as Pa. So Ha is doing the deciding and Pa is what is being decided on.

    Because the fixed algorithm of Ha aborts, it does not do a correct and complete emulation. Therefore "[Ha's] correct and complete x86 emulation of its input" does not exist, making the above statement nonsense.

    This has been told to you several times before and you have not attempted to explain why it is wrong. Therefore anyone that reads this is forced to conclude that you are unable to and therefore agree that your statement is nonsense.
    In other words Dennis and Richard are saying that no halt decider can >>>>>>>> ever correctly determine (in a finite number of steps) that
    Infinite_Loop() never halts.

    Infinite_Loop has previously shown to be irrelevant.

    Try again.

    Your claim is that it is impossible for a halt decider to correctly >>>>>> predict that its input would never halt unless this input is simulated >>>>>> forever and it never halts. I proved that you are wrong about this. >>>>>
    That's not what I said. I said that it's impossible for Ha(Pa,Pa) to do a correct and complete emulation of its input because it aborts, therefore its nonsense to predict what Ha(Pa,Pa)'s correct and complete emulation of its input will do.
    The halt deciders all correctly predict (in a finite number of steps)
    what their correct and complete emulation of their input would do on the >>>> basis that these inputs demonstrate non-halting behavior patterns that >>>> the halt deciders recognize.

    But in the case of Ha(Pa,Pa) it *doesn't* do a correct and complete emulation of its input, so it makes no sense to say what Ha's correct and complete emulation would be.


    If your nonsense reasoning applies to H(P,P) then when this same
    nonsense reasoning is applied to H0(Infinite_Loop) and
    H(Infinite_Recursion, 0x777) it is shown to be nonsense.

    Yes, it applies to those as well as none of them do a complete and correct simulation of their inputs. They happen to get the right answer because an *actual* correct and complete simulation of their inputs, i.e. UTM(Infinite_Loop) and UTM(Infinite_
    Recursion, 0x777) do not halt.

    The correct and complete simulation of the input to Ha(Pa,Pa),
    A simulating halt decider correctly predicts whether or not its correct
    and complete simulation of its input would ever reach the final state of
    this input.

    So you again just repeat your claim without explaining why I am wrong.

    Which means you agree that your claim is nonsense.

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.

    A halt decider computes the mapping from its inputs to an accept or
    reject state based on the actual behavior of these actual inputs.

    H(P,P) does do that therefore H(P,P)==0 is necessarily correct. Every
    rebuttal of a necessarily correct statement is necessarily incorrect.


    Many times its correct and complete simulation of its input

    Simply doesn't exist if the decider aborts

    would match
    the behavior of another different simulator simulating this same input.
    For any program H that might determine if programs halt, a
    "pathological"
    program P, called with some input, can pass its own source and its
    input to
    H and then specifically do the opposite of what H predicts P will
    do. No H
    can exist that handles this case.
    https://en.wikipedia.org/wiki/Halting_problem
    This is not the case when H and P have the above halting theorem
    pathological relationship to each other. In that case it is the actual
    behavior of the actual input that must be assessed.

    A halt decider must compute the mapping from its inputs to an accept or
    reject state on the basis of the actual behavior that is actually
    specified by these inputs.

    Which by the specification put forth by the halting problem:

    H(x,y)==1 if and only if x(y) halts, and
    H(x,y)==0 if and only if x(y) does not halt

    Is the behavior of x(y)

    Remember, the halting problem is asking the question "Does an *algorithm* exist that can determine if *any* arbitrary algorithm will halt given a particular input", and *if* such an algorithm exits, it MUST implement the above specification.


    P(P) reaches its "ret" instruction.
    From a purely software engineering perspective (anchored in the
    semantics of the x86 language)

    Ha must implement the specification it is stipulated to, which is

    H(x,y)==1 if and only if x(y) halts, and
    H(x,y)==0 if and only if x(y) does not halt

    it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input

    Does not exist if Ha aborts


    --
    Copyright 2022 Pete Olcott

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

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Richard Damon@21:1/5 to olcott on Wed Jul 6 21:35:01 2022
    XPost: comp.theory, sci.logic, sci.math

    On 7/6/22 9:32 AM, olcott wrote:
    On 7/6/2022 7:50 AM, Andy Walker wrote:
    On 05/07/2022 19:49, wij wrote:
    The idea of fork-simulation halting decider indeed looked much
    advanced and
    promising than the oral-based halting decider (POOH). Chance might be
    good
    refuting the HP.

         If by "refuting the HP" you mean "constructing a halt decider", then
    you have as much chance as refuting that 2+2 == 4 [in all cases, with the
    usual meanings of those words].  All the obfuscation of the last
    couple of
    decades does absolutely nothing to indicate any actual error in any of
    the
    several known proofs that no general halt decider can exist.  If you,
    or PO,
    ever did manage to produce an actual purported "H", then we already
    know how
    to construct an actual counterexample that refutes your, or his, claim.
    That's all anyone really needs to know.  We can sit back and wait however >> long it takes for an actual claimed "H" [whether in C or x86 code or as a
    TM] to appear, and then it is a matter of moments to produce a program
    and
    input that "H" fails with.

         If by "refuting the HP" you mean something else, then you need to >> explain further, as "refuting" in English applies to claims rather than
    to problems.


    I dare you to try to refute this.

    From a purely software engineering perspective (anchored in the
    semantics of the x86 language) it is proven that H(P,P) correctly
    predicts that its correct and complete x86 emulation of its input would
    never reach the "ret" instruction (final state) of this input.


    Since P(P) Halts when H(P,P) returns 0, this is obviously incorrect.

    Your dishonest dodge of saying the input to H(P,P) doesn't actually
    represent the program P(P) just says that H isn't actually meeting the definition of a Halt Decider.

    This is redoubled by the fact that the definition of P says P is asking
    H what P(P) does, so if H(P,P) doesn't mean that you also haven't
    constructed the correct P to match the proof you claim to refute.

    Thus, all you have proved is that your arguement is bogus and just based
    on misdefining things, in other words, you lie.

    void P(u32 x)
    {
      if (H(x, x))
        HERE: goto HERE;
      return;
    }

    int main()
    {
      Output("Input_Halts = ", H((u32)P, (u32)P));
    }

         For any program H that might determine if programs halt, a "pathological"
         program P, called with some input, can pass its own source and its input to
         H and then specifically do the opposite of what H predicts P will do. No H
         can exist that handles this case. https://en.wikipedia.org/wiki/Halting_problem

    H and P implement the above specified pathological relationship to each other:


    *Halting problem proofs refuted on the basis of software engineering* https://www.researchgate.net/publication/361701808_Halting_problem_proofs_refuted_on_the_basis_of_software_engineering




    Your argument: ALL LIES.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)