• Re: on ignoring the undecidable --- Liar Paradox disguised

    From olcott@polcott333@gmail.com to comp.theory on Sun Feb 8 19:06:11 2026
    From Newsgroup: comp.theory

    On 2/8/2026 6:48 PM, Ben Bacarisse wrote:
    dart200 <user7160@newsgrouper.org.invalid> writes:

    we should then consider a working thesis: no paradoxical machine is the
    simplest of their class of functionally equivalent machines.

    You don't define what a "paradoxical machine" is. Can you do so?

    I could take a guess, as it is a phrase commonly used by Usenet cranks
    to refer to a machine they think exists but does not. But I don't want
    to assume you've made the same mistake. Have you ever seen one? Can
    you show one? Can you define the term in a way that is close to being
    clear and unambiguous?

    consider the basic paradox form:

    deciderP(input) - decides if input has property P or !P
    machineP() - machine that has property P
    machine!P() - machine that has property !P

    // undecidable by deciderP for property P
    undP = () -> {
    if ( deciderP(undP) == TRUE )
    machine!P()
    else
    machineP()
    }

    This does not help. For example, let's assume decider37t(input)
    correctly decides (returns true) if the input machine executes more than
    37 state transitions (and false otherwise). What is paradoxical about und37t?

    Again, I could guess that you only want people to use this "basic
    paradox form" when deciderP does not exist, and so neither does undP;
    i.e. only when the conditions of Rice's theorem are satisfied. But you haven't said that, so I should let you clarify your definition.


    I proved that the Halting Problem is
    merely the Liar Paradox disguised 22 years ago https://groups.google.com/g/sci.logic/c/Hs78nMN6QZE/m/ID2rxwo__yQJ
    --
    Copyright 2026 Olcott<br><br>

    My 28 year goal has been to make <br>
    "true on the basis of meaning expressed in language"<br>
    reliably computable for the entire body of knowledge.<br><br>

    This required establishing a new foundation<br>
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Mon Feb 9 07:51:25 2026
    From Newsgroup: comp.theory

    On 2/8/26 8:06 PM, olcott wrote:
    On 2/8/2026 6:48 PM, Ben Bacarisse wrote:
    dart200 <user7160@newsgrouper.org.invalid> writes:

    we should then consider a working thesis: no paradoxical machine is the
    simplest of their class of functionally equivalent machines.

    You don't define what a "paradoxical machine" is.-a Can you do so?

    I could take a guess, as it is a phrase commonly used by Usenet cranks
    to refer to a machine they think exists but does not.-a But I don't want
    to assume you've made the same mistake.-a Have you ever seen one?-a Can
    you show one?-a Can you define the term in a way that is close to being
    clear and unambiguous?

    consider the basic paradox form:

    -a-a deciderP(input) - decides if input has property P or !P
    -a-a machineP()-a-a-a-a-a - machine that has property P
    -a-a machine!P()-a-a-a-a - machine that has property !P

    -a-a // undecidable by deciderP for property P
    -a-a undP = () -> {
    -a-a-a-a if ( deciderP(undP) == TRUE )
    -a-a-a-a-a-a machine!P()
    -a-a-a-a else
    -a-a-a-a-a-a machineP()
    -a-a }

    This does not help.-a For example, let's assume decider37t(input)
    correctly decides (returns true) if the input machine executes more than
    37 state transitions (and false otherwise).-a What is paradoxical about
    und37t?

    Again, I could guess that you only want people to use this "basic
    paradox form" when deciderP does not exist, and so neither does undP;
    i.e. only when the conditions of Rice's theorem are satisfied.-a But you
    haven't said that, so I should let you clarify your definition.


    I proved that the Halting Problem is
    merely the Liar Paradox disguised 22 years ago https://groups.google.com/g/sci.logic/c/Hs78nMN6QZE/m/ID2rxwo__yQJ


    No, you proved you don't understand what the Halting Problem *IS*

    The decider doesn't give its answer as an input the to machine under
    test, it returns that answer to its caller.

    Thus, all you do is prove your ignorance.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From olcott@polcott333@gmail.com to comp.theory on Mon Feb 9 08:01:06 2026
    From Newsgroup: comp.theory

    On 2/9/2026 6:51 AM, Richard Damon wrote:
    On 2/8/26 8:06 PM, olcott wrote:
    On 2/8/2026 6:48 PM, Ben Bacarisse wrote:
    dart200 <user7160@newsgrouper.org.invalid> writes:

    we should then consider a working thesis: no paradoxical machine is the >>>> simplest of their class of functionally equivalent machines.

    You don't define what a "paradoxical machine" is.-a Can you do so?

    I could take a guess, as it is a phrase commonly used by Usenet cranks
    to refer to a machine they think exists but does not.-a But I don't want >>> to assume you've made the same mistake.-a Have you ever seen one?-a Can
    you show one?-a Can you define the term in a way that is close to being
    clear and unambiguous?

    consider the basic paradox form:

    -a-a deciderP(input) - decides if input has property P or !P
    -a-a machineP()-a-a-a-a-a - machine that has property P
    -a-a machine!P()-a-a-a-a - machine that has property !P

    -a-a // undecidable by deciderP for property P
    -a-a undP = () -> {
    -a-a-a-a if ( deciderP(undP) == TRUE )
    -a-a-a-a-a-a machine!P()
    -a-a-a-a else
    -a-a-a-a-a-a machineP()
    -a-a }

    This does not help.-a For example, let's assume decider37t(input)
    correctly decides (returns true) if the input machine executes more than >>> 37 state transitions (and false otherwise).-a What is paradoxical about
    und37t?

    Again, I could guess that you only want people to use this "basic
    paradox form" when deciderP does not exist, and so neither does undP;
    i.e. only when the conditions of Rice's theorem are satisfied.-a But you >>> haven't said that, so I should let you clarify your definition.


    I proved that the Halting Problem is
    merely the Liar Paradox disguised 22 years ago
    https://groups.google.com/g/sci.logic/c/Hs78nMN6QZE/m/ID2rxwo__yQJ


    No, you proved you don't understand what the Halting Problem *IS*

    The decider doesn't give its answer as an input the to machine under
    test, it returns that answer to its caller.

    Thus, all you do is prove your ignorance.

    22 years ago I made you the decider and you
    could not decide because the HP <is> the LP.
    --
    Copyright 2026 Olcott<br><br>

    My 28 year goal has been to make <br>
    "true on the basis of meaning expressed in language"<br>
    reliably computable for the entire body of knowledge.<br><br>

    This required establishing a new foundation<br>
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Mon Feb 9 22:47:29 2026
    From Newsgroup: comp.theory

    On 2/9/26 9:01 AM, olcott wrote:
    On 2/9/2026 6:51 AM, Richard Damon wrote:
    On 2/8/26 8:06 PM, olcott wrote:
    On 2/8/2026 6:48 PM, Ben Bacarisse wrote:
    dart200 <user7160@newsgrouper.org.invalid> writes:

    we should then consider a working thesis: no paradoxical machine is >>>>> the
    simplest of their class of functionally equivalent machines.

    You don't define what a "paradoxical machine" is.-a Can you do so?

    I could take a guess, as it is a phrase commonly used by Usenet cranks >>>> to refer to a machine they think exists but does not.-a But I don't want >>>> to assume you've made the same mistake.-a Have you ever seen one?-a Can >>>> you show one?-a Can you define the term in a way that is close to being >>>> clear and unambiguous?

    consider the basic paradox form:

    -a-a deciderP(input) - decides if input has property P or !P
    -a-a machineP()-a-a-a-a-a - machine that has property P
    -a-a machine!P()-a-a-a-a - machine that has property !P

    -a-a // undecidable by deciderP for property P
    -a-a undP = () -> {
    -a-a-a-a if ( deciderP(undP) == TRUE )
    -a-a-a-a-a-a machine!P()
    -a-a-a-a else
    -a-a-a-a-a-a machineP()
    -a-a }

    This does not help.-a For example, let's assume decider37t(input)
    correctly decides (returns true) if the input machine executes more
    than
    37 state transitions (and false otherwise).-a What is paradoxical about >>>> und37t?

    Again, I could guess that you only want people to use this "basic
    paradox form" when deciderP does not exist, and so neither does undP;
    i.e. only when the conditions of Rice's theorem are satisfied.-a But you >>>> haven't said that, so I should let you clarify your definition.


    I proved that the Halting Problem is
    merely the Liar Paradox disguised 22 years ago
    https://groups.google.com/g/sci.logic/c/Hs78nMN6QZE/m/ID2rxwo__yQJ


    No, you proved you don't understand what the Halting Problem *IS*

    The decider doesn't give its answer as an input the to machine under
    test, it returns that answer to its caller.

    Thus, all you do is prove your ignorance.

    22 years ago I made you the decider and you
    could not decide because the HP <is> the LP.



    Nope.

    YOU are the Liar, and can't handle the paradox because you are just a
    stupid liar.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From olcott@polcott333@gmail.com to comp.theory,sci.logic,sci.math on Thu Feb 12 11:41:55 2026
    From Newsgroup: comp.theory

    On 2/12/2026 11:23 AM, Tristan Wibberley wrote:
    On 11/02/2026 21:53, Alan Mackenzie wrote:
    Mathematicians have proven that many decision problems can not be
    answered, all the nonsense about "idiosyncrasies of self-referential
    logic" notwithstanding.


    It's not at all clear to me that those unanswerables are properly
    classified as "decision problem" unless one uses an auto-explication (my
    term for when a term is both an explicatum and explicandum of an explication). Carnap's definition of explication excludes such an act
    (though I don't know if he'd picked up the bad habit of using "decision problem" as an explicatum).

    I think Carnap would have admitted "L-decision problem" as an explicatum
    of the explicandum "decision problem". The nonsense act of calling pathologically self-referential problems as "L-decision problems" would
    be obvious because one does not have a problem of choosing between only classifications "true" and "false" when one has merely been fooled into thinking those are candidates without a whole heap of others beside.

    I hereby indulge myself with some old-timey assertive logistic
    philosophy, you might call it a strawman, something to ponder and burn down:

    We can understand the fallacy by making explicit the implicit false assumption: "the sentence after the conjunctive connector following can
    be assigned no valuation but 'true' or 'false' AND blah-blah". That is
    the cultural synergy covertly induced in the ponderer by a poetic form
    of expression ("proposition" the explicatum, not the explicandum, it's another auto-explication) but it's not /well/ formalised in that it's a
    mess of massive description and wonderment. I think usage of AND gives
    us falsity for pathologically self-referential 'blah-blah' if we have
    the right type-system but other connectives give us other, non-truth, classifications. A connective that means the implication of neither
    truth nor falsity is also available. Of course, the ponderer has the inducement in the form of a volition to be disobedient and choose to
    react in a variety of unassertive ways.


    A proof is a proof.

    Tautology.


    *My pair of axioms is confirmed by peer reviewed papers*
    reCx (Provable(T, x) rco Meaningful(T, x)) --- (Schroeder-Heister 2024)
    reCx (Provable(x) rcA True(x)) --- Anchored in (Prawitz, 2012)

    What is the appropriate notion of truth for sentences whose meanings are understood in epistemic terms such as proof or ground for an assertion?
    It seems that the truth of such sentences has to be identified with the existence of proofs or grounds...
    Prawitz, D. (2012). Truth as an Epistemic Notion. Topoi, 31(1), 9rCo16 https://doi.org/10.1007/s11245-011-9107-6

    1.2 Inferentialism, intuitionism, anti-realism
    Proof-theoretic semantics is inherently inferential, as it is
    inferential activity which manifests itself in proofs. It thus belongs
    to inferentialism (a term coined by Brandom, see his 1994; 2000)
    according to which inferences and the rules of inference establish the
    meaning of expressions...
    Schroeder-Heister, Peter, 2024 "Proof-Theoretic Semantics" https://plato.stanford.edu/entries/proof-theoretic-semantics/#InfeIntuAntiReal --
    Copyright 2026 Olcott<br><br>

    My 28 year goal has been to make <br>
    "true on the basis of meaning expressed in language"<br>
    reliably computable for the entire body of knowledge.<br><br>

    This required establishing a new foundation<br>
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From olcott@polcott333@gmail.com to comp.theory on Thu Feb 12 19:59:19 2026
    From Newsgroup: comp.theory

    On 2/12/2026 7:39 PM, Tristan Wibberley wrote:
    On 12/02/2026 20:02, Alan Mackenzie wrote:
    [ Followup-To: set ]

    In comp.theory Tristan Wibberley
    <tristan.wibberley+netnews2@alumni.manchester.ac.uk> wrote:
    On 11/02/2026 21:53, Alan Mackenzie wrote:
    Mathematicians have proven that many decision problems can not be
    answered, all the nonsense about "idiosyncrasies of self-referential
    logic" notwithstanding.

    It's not at all clear to me that those unanswerables are properly
    classified as "decision problem" unless one uses an auto-explication (my >>> term for when a term is both an explicatum and explicandum of an
    explication). Carnap's definition of explication excludes such an act
    (though I don't know if he'd picked up the bad habit of using "decision
    problem" as an explicatum).

    My understanding of a "decision problem" is one whose solution is a
    machine which, in finite time, can correctly classify any machine into
    one of two categories.

    Then a problem without a solution is not a decision problem; I think
    that's the right way to think about it. I think Olcott thinks about it
    that way too, at least for days at a time. Conversations here manage to
    make me slip between modes sometimes.


    In this sense there is no solution to the halting problem.


    In the same sense that no one will ever figure out
    if this sentence is true or false: "What time is is?"
    thus forever blocking truth predicate.

    And by your partial explication above, the halting problem is therefore
    not a decision problem. I don't think your explication will stick still
    for more than half a sentence, though; at any moment I think you or
    someone else studious will say that there /is/ a solution to the halting problem, the solution they will moot is that there is no universal halt-decider.


    --
    Copyright 2026 Olcott<br><br>

    My 28 year goal has been to make <br>
    "true on the basis of meaning expressed in language"<br>
    reliably computable for the entire body of knowledge.<br><br>

    This required establishing a new foundation<br>
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From user7160@user7160@newsgrouper.org.invalid to comp.theory on Tue Feb 17 07:02:39 2026
    From Newsgroup: comp.theory

    This message was cancelled from within Thunderbird.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Alan Mackenzie@acm@muc.de to comp.theory on Thu Feb 19 14:43:28 2026
    From Newsgroup: comp.theory

    Ben Bacarisse <ben@bsb.me.uk> wrote:

    [ .... ]

    The Church-Turing thesis is clearly not the kind of thing that is
    provable. That's why it's not called a conjecture.

    I don't understand that bit. What is unprovable about the Church-Turing thesis? I think it hypothesises that there is no computing machine more powerful than a turing machine.

    What would prevent a proof along the lines of supposing the existence of
    such a more powerful machine, then proving it was actually equivalent to
    some turing machine?

    [ .... ]

    --
    Ben.

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ben Bacarisse@ben@bsb.me.uk to comp.theory on Fri Feb 20 00:24:06 2026
    From Newsgroup: comp.theory

    Alan Mackenzie <acm@muc.de> writes:

    Ben Bacarisse <ben@bsb.me.uk> wrote:

    [ .... ]

    The Church-Turing thesis is clearly not the kind of thing that is
    provable. That's why it's not called a conjecture.

    I don't understand that bit. What is unprovable about the Church-Turing thesis? I think it hypothesises that there is no computing machine more powerful than a turing machine.

    There are, purely theoretical, models of computation (my preferred
    phrase) that are more powerful than Turing machines but they are not
    considered "effective". The thesis is about what effectively computable
    means and this is where the problem lies. It's not a well-defined
    concept, though almost everyone just /assumes/ it means "what TMs (and so
    on) can do".

    What would prevent a proof along the lines of supposing the existence of
    such a more powerful machine, then proving it was actually equivalent to
    some turing machine?

    Perhaps it would have been better to say that one can't imagine what
    such a proof could look like. Can you? The "more powerful" bit is
    easy. One would assume that the more powerful model can compute at
    least one function that is not TM computable. But how would the notion
    that it is none-the less "effective" be specified? And what form could
    the equivalence proof take, given that we can assume nothing about model
    other that the fact that is it more powerful and yet effective?

    It's not something I've given much thought to so I'd be interested if
    you can go further with the notion.
    --
    Ben.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Alan Mackenzie@acm@muc.de to comp.theory on Fri Feb 20 16:07:27 2026
    From Newsgroup: comp.theory

    Ben Bacarisse <ben@bsb.me.uk> wrote:
    Alan Mackenzie <acm@muc.de> writes:

    Ben Bacarisse <ben@bsb.me.uk> wrote:

    [ .... ]

    The Church-Turing thesis is clearly not the kind of thing that is
    provable. That's why it's not called a conjecture.

    I don't understand that bit. What is unprovable about the Church-Turing
    thesis? I think it hypothesises that there is no computing machine more
    powerful than a turing machine.

    There are, purely theoretical, models of computation (my preferred
    phrase) that are more powerful than Turing machines but they are not considered "effective". The thesis is about what effectively computable means and this is where the problem lies. It's not a well-defined
    concept, though almost everyone just /assumes/ it means "what TMs (and so
    on) can do".

    I think I'm beginning to see the problem. These more powerful models of computation presumably lack the "finite structure" of a turing machine -
    the finite number of states, of possible symbols on the tape, and the discreteness of the tape movements.

    Maybe analogue devices (slide rules, differential analysers, etc.) come
    into this category. Though these could not duplicate the effect of a
    turing machine, they do something altogether different.

    What would prevent a proof along the lines of supposing the existence of
    such a more powerful machine, then proving it was actually equivalent to
    some turing machine?

    Perhaps it would have been better to say that one can't imagine what
    such a proof could look like. Can you?

    No. I'm pretty confused about the whole question.

    The "more powerful" bit is easy. One would assume that the more
    powerful model can compute at least one function that is not TM
    computable.

    There are only a countable number of turing machines, but an uncountable
    number of functions. So that "at least one" would probably be an
    uncountably infinite number.

    But how would the notion that it is none-the less "effective" be
    specified?

    I think it needs to be a machine of some sort, in the sense of being a
    finite collection of rods, gears, states, tapes, whatever ... But I
    can't picture any such device which wouldn't be equivalent to a turing
    machine.

    And what form could the equivalence proof take, given that we can
    assume nothing about model other that the fact that is it more
    powerful and yet effective?

    Maybe there could be a proof that any machine worthy of the description
    would be equivalent to a turing machine. This would need to formalise
    exactly what a "machine" is. Maybe this has been done already.

    It's not something I've given much thought to so I'd be interested if
    you can go further with the notion.

    I'll see if I can come up with something more coherent after some more
    thought.

    --
    Ben.
    --
    Alan Mackenzie (Nuremberg, Germany).

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ben Bacarisse@ben@bsb.me.uk to comp.theory on Fri Feb 20 18:11:37 2026
    From Newsgroup: comp.theory

    Alan Mackenzie <acm@muc.de> writes:

    Ben Bacarisse <ben@bsb.me.uk> wrote:
    Alan Mackenzie <acm@muc.de> writes:

    Ben Bacarisse <ben@bsb.me.uk> wrote:

    [ .... ]

    The Church-Turing thesis is clearly not the kind of thing that is
    provable. That's why it's not called a conjecture.

    I don't understand that bit. What is unprovable about the Church-Turing >>> thesis? I think it hypothesises that there is no computing machine more >>> powerful than a turing machine.

    There are, purely theoretical, models of computation (my preferred
    phrase) that are more powerful than Turing machines but they are not
    considered "effective". The thesis is about what effectively computable
    means and this is where the problem lies. It's not a well-defined
    concept, though almost everyone just /assumes/ it means "what TMs (and so
    on) can do".

    I think I'm beginning to see the problem. These more powerful models of computation presumably lack the "finite structure" of a turing machine -
    the finite number of states, of possible symbols on the tape, and the discreteness of the tape movements.

    The most comment extensions are orcale machines that are TMs with an
    extra tape containing conventionally uncomputable data, most commonly
    the tape can be accessed to determine the halting of a given TM. They
    are dreamt up simply to show that such machines also have a halting
    problem of their own but they clearly don't capture the notion of
    something being "effectively computable".

    Maybe analogue devices (slide rules, differential analysers, etc.) come
    into this category. Though these could not duplicate the effect of a
    turing machine, they do something altogether different.

    What would prevent a proof along the lines of supposing the existence of >>> such a more powerful machine, then proving it was actually equivalent to >>> some turing machine?

    Perhaps it would have been better to say that one can't imagine what
    such a proof could look like. Can you?

    No. I'm pretty confused about the whole question.

    The "more powerful" bit is easy. One would assume that the more
    powerful model can compute at least one function that is not TM
    computable.

    There are only a countable number of turing machines, but an uncountable number of functions. So that "at least one" would probably be an
    uncountably infinite number.

    But how would the notion that it is none-the less "effective" be
    specified?

    I think it needs to be a machine of some sort, in the sense of being a
    finite collection of rods, gears, states, tapes, whatever ... But I
    can't picture any such device which wouldn't be equivalent to a turing machine.

    The proof would have to assume one exists but nothing else about it all
    other than the fact that it's computations are what we'd call effective computations. Any other assumptions about it would render the proof
    invalid. Well, it would mean that we are not proving the Church Turing
    thesis but something about the specific class of machines we have not
    dreamt up.

    And what form could the equivalence proof take, given that we can
    assume nothing about model other that the fact that is it more
    powerful and yet effective?

    Maybe there could be a proof that any machine worthy of the description
    would be equivalent to a turing machine. This would need to formalise exactly what a "machine" is. Maybe this has been done already.

    That's exactly the trouble. How do we formalise what we mean without
    either being so specific that it's just another particular model or
    being so vague that we can't make any deduction from the formalisation.

    Turing invented TMs to capture the notion of what is computable, and it
    is so compelling because it seems to capture the idea very well. Every
    one "gets it" and accepts the idea that this is what we mean by an
    algorithm. Church's model -- the lambda calculus -- does not appear to
    capture it at all. If you described it to someone bright but ignorant
    of computing they would never say "ah, yes, everything I think of as
    being a computation is clearly equivalent to some lambda form"!

    It's a lovely irony that it's the lambda calculus that has turned out to
    be useful in practice, whereas Turing machines are used only in theory
    courses.

    It's not something I've given much thought to so I'd be interested if
    you can go further with the notion.

    I'll see if I can come up with something more coherent after some more thought.

    --
    Ben.
    --
    Ben.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Tristan Wibberley@tristan.wibberley+netnews2@alumni.manchester.ac.uk to comp.theory on Tue Feb 24 20:10:59 2026
    From Newsgroup: comp.theory

    On 20/02/2026 00:24, Ben Bacarisse wrote:
    Alan Mackenzie <acm@muc.de> writes:

    Ben Bacarisse <ben@bsb.me.uk> wrote:

    [ .... ]

    The Church-Turing thesis is clearly not the kind of thing that is
    provable. That's why it's not called a conjecture.

    I don't understand that bit. What is unprovable about the Church-Turing
    thesis? I think it hypothesises that there is no computing machine more
    powerful than a turing machine.

    There are, purely theoretical, models of computation (my preferred
    phrase) that are more powerful than Turing machines but they are not considered "effective".

    Perhaps a new explicatum would help: "solvation"

    The semi-religious nutjobs can go totally insane at each other with that
    one in the conversation... It'll be fun.

    Computation is for solving things, in general we could refer to
    solvation, either by computation, by oracular inquiry, by divination, by miraculous transmutation of problem statement to solution, by gradient
    descent of physical structures (decay). Perhaps things that modern
    computers do that use entropy to choose a path that approaches a
    solution could be called solvation by compudecay.
    --
    Tristan Wibberley

    The message body is Copyright (C) 2026 Tristan Wibberley except
    citations and quotations noted. All Rights Reserved except that you may,
    of course, cite it academically giving credit to me, distribute it
    verbatim as part of a usenet system or its archives, and use it to
    promote my greatness and general superiority without misrepresentation
    of my opinions other than my opinion of my greatness and general
    superiority which you _may_ misrepresent. You definitely MAY NOT train
    any production AI system with it but you may train experimental AI that
    will only be used for evaluation of the AI methods it implements.

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Tristan Wibberley@tristan.wibberley+netnews2@alumni.manchester.ac.uk to comp.theory on Tue Feb 24 20:15:17 2026
    From Newsgroup: comp.theory

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known. He worked in defence
    research so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.
    --
    Tristan Wibberley

    The message body is Copyright (C) 2026 Tristan Wibberley except
    citations and quotations noted. All Rights Reserved except that you may,
    of course, cite it academically giving credit to me, distribute it
    verbatim as part of a usenet system or its archives, and use it to
    promote my greatness and general superiority without misrepresentation
    of my opinions other than my opinion of my greatness and general
    superiority which you _may_ misrepresent. You definitely MAY NOT train
    any production AI system with it but you may train experimental AI that
    will only be used for evaluation of the AI methods it implements.

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Tue Feb 24 19:01:33 2026
    From Newsgroup: comp.theory

    On 2/24/26 3:15 PM, Tristan Wibberley wrote:
    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known. He worked in defence
    research so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.


    Maybe you should read more of his writing where he talks about what he
    is showing.

    Turing was very much interested in the science of logic and the idea of
    using computing.

    Note, "Enemy Computing Infrastructure" didn't really exist at the time,
    as Computers as we know it didn't exist yet.

    I guess you don't understand that part of history.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Ben Bacarisse@ben@bsb.me.uk to comp.theory on Wed Feb 25 00:09:56 2026
    From Newsgroup: comp.theory

    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear. In 1928, Hilbert and Ackermann posed
    the question of whether first-order logic was decidable -- was there an algorithm that can determine if a given statement is universally valid.
    (Given other results this is equivalent to deciding, algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church who had
    turned his attention to this as yet unsolved problem. At the time, most mathematicians thought the answer would be "yes". Of course the first
    step is to capture the notion of an algorithm or process. Church came
    up with the lambda calculus, and Turing the abstract machine that not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be
    automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no. He was just a PhD student interested in formal
    logic and the intriguing unsolved problems of the time, one of which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.
    --
    Ben.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Tue Feb 24 16:39:02 2026
    From Newsgroup: comp.theory

    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk> writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear. In 1928, Hilbert and Ackermann posed
    the question of whether first-order logic was decidable -- was there an

    and they all gave up on essentially meaninglessly garbage like

    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all
    twisted up /in the same way/ because of failure to appropriately handle
    a lil' self-referential contradiction

    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    algorithm that can determine if a given statement is universally valid. (Given other results this is equivalent to deciding, algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church who had
    turned his attention to this as yet unsolved problem. At the time, most mathematicians thought the answer would be "yes". Of course the first
    step is to capture the notion of an algorithm or process. Church came
    up with the lambda calculus, and Turing the abstract machine that not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no. He was just a PhD student interested in formal
    logic and the intriguing unsolved problems of the time, one of which was about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.

    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Tue Feb 24 21:13:02 2026
    From Newsgroup: comp.theory

    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann posed
    the question of whether first-order logic was decidable -- was there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all twisted up /in the same way/ because of failure to appropriately handle
    a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.


    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think what
    it would do).


    algorithm that can determine if a given statement is universally valid.
    (Given other results this is equivalent to deciding, algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church who had
    turned his attention to this as yet unsolved problem.-a At the time, most
    mathematicians thought the answer would be "yes".-a Of course the first
    step is to capture the notion of an algorithm or process.-a Church came
    up with the lambda calculus, and Turing the abstract machine that not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be
    automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal
    logic and the intriguing unsolved problems of the time, one of which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.



    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Tue Feb 24 18:47:16 2026
    From Newsgroup: comp.theory

    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann posed >>> the question of whether first-order logic was decidable -- was there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all
    twisted up /in the same way/ because of failure to appropriately
    handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...



    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think what
    it would do).


    algorithm that can determine if a given statement is universally valid.
    (Given other results this is equivalent to deciding, algorithmically, if >>> a given statement is provable.)

    At the time of his now famous paper he was working with Church who had
    turned his attention to this as yet unsolved problem.-a At the time, most >>> mathematicians thought the answer would be "yes".-a Of course the first
    step is to capture the notion of an algorithm or process.-a Church came
    up with the lambda calculus, and Turing the abstract machine that not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be
    automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal
    logic and the intriguing unsolved problems of the time, one of which was >>> about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.



    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From user7160@user7160@newsgrouper.org.invalid to comp.theory on Wed Feb 25 03:01:43 2026
    From Newsgroup: comp.theory

    This message was cancelled from within Thunderbird.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Tue Feb 24 19:06:00 2026
    From Newsgroup: comp.theory

    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann posed >>> the question of whether first-order logic was decidable -- was there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all
    twisted up /in the same way/ because of failure to appropriately
    handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N, which turing explicitly notes, and
    is the exact point where D fails to make an appropriate decision

    why are you so fking dumb richard? is it going to take a third time
    quoting that at you???



    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think what
    it would do).


    algorithm that can determine if a given statement is universally valid.
    (Given other results this is equivalent to deciding, algorithmically, if >>> a given statement is provable.)

    At the time of his now famous paper he was working with Church who had
    turned his attention to this as yet unsolved problem.-a At the time, most >>> mathematicians thought the answer would be "yes".-a Of course the first
    step is to capture the notion of an algorithm or process.-a Church came
    up with the lambda calculus, and Turing the abstract machine that not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be
    automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal
    logic and the intriguing unsolved problems of the time, one of which was >>> about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.



    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Tue Feb 24 22:52:32 2026
    From Newsgroup: comp.theory

    On 2/24/26 9:47 PM, dart200 wrote:
    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk> >>>> writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann
    posed
    the question of whether first-order logic was decidable -- was there an >>>
    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all
    twisted up /in the same way/ because of failure to appropriately
    handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...

    But it doesn't have a reference to that. it computed that.

    You don't seem to understand what a "reference" is.

    Do you think that a compiler can't compile the code for itself?


    Yes, a decider that decides on all inputs (as the decision problem
    requires) needs to be able to decide about the "value" that it
    represents itself.

    In part, the undecidability comes out of the fact that the machinery of
    the system IS powerful enough that we can convert the machines into
    values that can be their inputs.

    Since "Turing Complete" machines can do this, any system that can't must
    be less powerful than "Turing Complete".




    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think
    what it would do).


    algorithm that can determine if a given statement is universally valid. >>>> (Given other results this is equivalent to deciding,
    algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church who had >>>> turned his attention to this as yet unsolved problem.-a At the time,
    most
    mathematicians thought the answer would be "yes".-a Of course the first >>>> step is to capture the notion of an algorithm or process.-a Church came >>>> up with the lambda calculus, and Turing the abstract machine that not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be
    automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal
    logic and the intriguing unsolved problems of the time, one of which
    was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources.






    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Tue Feb 24 20:02:49 2026
    From Newsgroup: comp.theory

    On 2/24/26 7:52 PM, Richard Damon wrote:
    On 2/24/26 9:47 PM, dart200 wrote:
    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk> >>>>> writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann >>>>> posed
    the question of whether first-order logic was decidable -- was
    there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties
    all twisted up /in the same way/ because of failure to appropriately
    handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...

    But it doesn't have a reference to that. it computed that.

    You don't seem to understand what a "reference" is.

    it has an addressable copy of it's source code on the tape, encoded into
    a description number, that it tests

    that's the only kind of self-reference that exists in turing machines,
    and it always involves some steps in the computation to setup, unless
    it's directly set as the input ... but either way it's a reference to
    itself, and the reference to itself is keystone in undecidability proofs


    Do you think that a compiler can't compile the code for itself?


    Yes, a decider that decides on all inputs (as the decision problem
    requires) needs to be able to decide about the "value" that it
    represents itself.

    In part, the undecidability comes out of the fact that the machinery of
    the system IS powerful enough that we can convert the machines into
    values that can be their inputs.

    Since "Turing Complete" machines can do this, any system that can't must
    be less powerful than "Turing Complete".




    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think
    what it would do).


    algorithm that can determine if a given statement is universally
    valid.
    (Given other results this is equivalent to deciding,
    algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church who had >>>>> turned his attention to this as yet unsolved problem.-a At the time, >>>>> most
    mathematicians thought the answer would be "yes".-a Of course the first >>>>> step is to capture the notion of an algorithm or process.-a Church came >>>>> up with the lambda calculus, and Turing the abstract machine that not >>>>> bears his name.

    However, the idea that mathematics and specifically proofs, might be >>>>> automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal >>>>> logic and the intriguing unsolved problems of the time, one of
    which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources. >>>>>





    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Tue Feb 24 23:29:16 2026
    From Newsgroup: comp.theory

    On 2/24/26 11:02 PM, dart200 wrote:
    On 2/24/26 7:52 PM, Richard Damon wrote:
    On 2/24/26 9:47 PM, dart200 wrote:
    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley
    <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable >>>>>>>
    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann >>>>>> posed
    the question of whether first-order logic was decidable -- was
    there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties
    all twisted up /in the same way/ because of failure to
    appropriately handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing
    Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...

    But it doesn't have a reference to that. it computed that.

    You don't seem to understand what a "reference" is.

    it has an addressable copy of it's source code on the tape, encoded into
    a description number, that it tests

    that's the only kind of self-reference that exists in turing machines,
    and it always involves some steps in the computation to setup, unless
    it's directly set as the input ... but either way it's a reference to itself, and the reference to itself is keystone in undecidability proofs


    Which isn't a "Reference" in the meaning of the word.

    Just shows that Turing Machines CAN'T have "self-references" as defined
    in logic.

    It is processing an input that just happens to match a description of
    itself.

    Since that is a legal input, no foul.

    In other words, to try to prohibit this input, means you can't be as
    powerful in computing as a Turing Machine.

    Thus, this "problem" is ESSENTIAL to the nature of the machines, and
    can't be just "fixed".


    Do you think that a compiler can't compile the code for itself?


    Yes, a decider that decides on all inputs (as the decision problem
    requires) needs to be able to decide about the "value" that it
    represents itself.

    In part, the undecidability comes out of the fact that the machinery
    of the system IS powerful enough that we can convert the machines into
    values that can be their inputs.

    Since "Turing Complete" machines can do this, any system that can't
    must be less powerful than "Turing Complete".




    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to
    kill myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think
    what it would do).


    algorithm that can determine if a given statement is universally
    valid.
    (Given other results this is equivalent to deciding,
    algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church who >>>>>> had
    turned his attention to this as yet unsolved problem.-a At the
    time, most
    mathematicians thought the answer would be "yes".-a Of course the >>>>>> first
    step is to capture the notion of an algorithm or process.-a Church >>>>>> came
    up with the lambda calculus, and Turing the abstract machine that not >>>>>> bears his name.

    However, the idea that mathematics and specifically proofs, might be >>>>>> automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal >>>>>> logic and the intriguing unsolved problems of the time, one of
    which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research resources. >>>>>>








    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Tue Feb 24 21:42:16 2026
    From Newsgroup: comp.theory

    On 2/24/26 8:29 PM, Richard Damon wrote:
    On 2/24/26 11:02 PM, dart200 wrote:
    On 2/24/26 7:52 PM, Richard Damon wrote:
    On 2/24/26 9:47 PM, dart200 wrote:
    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley
    <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable >>>>>>>>
    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and
    Ackermann posed
    the question of whether first-order logic was decidable -- was
    there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties >>>>>> all twisted up /in the same way/ because of failure to
    appropriately handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing
    Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...

    But it doesn't have a reference to that. it computed that.

    You don't seem to understand what a "reference" is.

    it has an addressable copy of it's source code on the tape, encoded
    into a description number, that it tests

    that's the only kind of self-reference that exists in turing machines,
    and it always involves some steps in the computation to setup, unless
    it's directly set as the input ... but either way it's a reference to
    itself, and the reference to itself is keystone in undecidability proofs


    Which isn't a "Reference" in the meaning of the word.

    not worthy of my time to debate semantics over something making up
    conflicts that need not exist

    it's a self-reference for anyone but someone bent on being contrarian
    for the sake of it, and i will keep referring to it like regardless of
    ur objections


    Just shows that Turing Machines CAN'T have "self-references" as defined
    in logic.

    It is processing an input that just happens to match a description of itself.

    Since that is a legal input, no foul.

    In other words, to try to prohibit this input, means you can't be as powerful in computing as a Turing Machine.

    Thus, this "problem" is ESSENTIAL to the nature of the machines, and
    can't be just "fixed".


    Do you think that a compiler can't compile the code for itself?


    Yes, a decider that decides on all inputs (as the decision problem
    requires) needs to be able to decide about the "value" that it
    represents itself.

    In part, the undecidability comes out of the fact that the machinery
    of the system IS powerful enough that we can convert the machines
    into values that can be their inputs.

    Since "Turing Complete" machines can do this, any system that can't
    must be less powerful than "Turing Complete".




    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to
    kill myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think >>>>> what it would do).


    algorithm that can determine if a given statement is universally >>>>>>> valid.
    (Given other results this is equivalent to deciding,
    algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church
    who had
    turned his attention to this as yet unsolved problem.-a At the
    time, most
    mathematicians thought the answer would be "yes".-a Of course the >>>>>>> first
    step is to capture the notion of an algorithm or process.-a Church >>>>>>> came
    up with the lambda calculus, and Turing the abstract machine that >>>>>>> not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be >>>>>>> automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal >>>>>>> logic and the intriguing unsolved problems of the time, one of
    which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research
    resources.









    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From user7160@user7160@newsgrouper.org.invalid to comp.theory on Wed Feb 25 08:35:57 2026
    From Newsgroup: comp.theory

    This message was cancelled from within Thunderbird.
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Wed Feb 25 00:36:47 2026
    From Newsgroup: comp.theory

    On 2/24/26 8:29 PM, Richard Damon wrote:
    On 2/24/26 11:02 PM, dart200 wrote:
    On 2/24/26 7:52 PM, Richard Damon wrote:
    On 2/24/26 9:47 PM, dart200 wrote:
    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley
    <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable >>>>>>>>
    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and
    Ackermann posed
    the question of whether first-order logic was decidable -- was
    there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties >>>>>> all twisted up /in the same way/ because of failure to
    appropriately handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing
    Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...

    But it doesn't have a reference to that. it computed that.

    You don't seem to understand what a "reference" is.

    it has an addressable copy of it's source code on the tape, encoded
    into a description number, that it tests

    that's the only kind of self-reference that exists in turing machines,
    and it always involves some steps in the computation to setup, unless
    it's directly set as the input ... but either way it's a reference to
    itself, and the reference to itself is keystone in undecidability proofs


    Which isn't a "Reference" in the meaning of the word.

    not worthy of my time to debate semantics with someone making up
    conflicts that need not exist

    it's adequately a self-reference for anyone but one bent on being
    contrarian for the sake of it, and i will keep referring to it like that regardless of ur objections


    Just shows that Turing Machines CAN'T have "self-references" as defined
    in logic.

    It is processing an input that just happens to match a description of itself.

    Since that is a legal input, no foul.

    In other words, to try to prohibit this input, means you can't be as powerful in computing as a Turing Machine.

    Thus, this "problem" is ESSENTIAL to the nature of the machines, and
    can't be just "fixed".


    Do you think that a compiler can't compile the code for itself?


    Yes, a decider that decides on all inputs (as the decision problem
    requires) needs to be able to decide about the "value" that it
    represents itself.

    In part, the undecidability comes out of the fact that the machinery
    of the system IS powerful enough that we can convert the machines
    into values that can be their inputs.

    Since "Turing Complete" machines can do this, any system that can't
    must be less powerful than "Turing Complete".




    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to
    kill myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think >>>>> what it would do).


    algorithm that can determine if a given statement is universally >>>>>>> valid.
    (Given other results this is equivalent to deciding,
    algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church
    who had
    turned his attention to this as yet unsolved problem.-a At the
    time, most
    mathematicians thought the answer would be "yes".-a Of course the >>>>>>> first
    step is to capture the notion of an algorithm or process.-a Church >>>>>>> came
    up with the lambda calculus, and Turing the abstract machine that >>>>>>> not
    bears his name.

    However, the idea that mathematics and specifically proofs, might be >>>>>>> automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in formal >>>>>>> logic and the intriguing unsolved problems of the time, one of
    which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability
    of enemy computing infrastructure to help direct research
    resources.









    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Richard Damon@Richard@Damon-Family.org to comp.theory on Wed Feb 25 07:04:47 2026
    From Newsgroup: comp.theory

    On 2/25/26 12:42 AM, dart200 wrote:
    On 2/24/26 8:29 PM, Richard Damon wrote:
    On 2/24/26 11:02 PM, dart200 wrote:
    On 2/24/26 7:52 PM, Richard Damon wrote:
    On 2/24/26 9:47 PM, dart200 wrote:
    On 2/24/26 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley
    <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable >>>>>>>>>
    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and
    Ackermann posed
    the question of whether first-order logic was decidable -- was >>>>>>>> there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties >>>>>>> all twisted up /in the same way/ because of failure to
    appropriately handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing
    Formulation.

    i've already quoted this at you

    /Now let K be the D.N of H. What does H do in the K-th section of
    its motion? It must test whether K is satisfactory/

    H is literally testing it's own D.N...

    But it doesn't have a reference to that. it computed that.

    You don't seem to understand what a "reference" is.

    it has an addressable copy of it's source code on the tape, encoded
    into a description number, that it tests

    that's the only kind of self-reference that exists in turing
    machines, and it always involves some steps in the computation to
    setup, unless it's directly set as the input ... but either way it's
    a reference to itself, and the reference to itself is keystone in
    undecidability proofs


    Which isn't a "Reference" in the meaning of the word.

    not worthy of my time to debate semantics over something making up
    conflicts that need not exist

    it's a self-reference for anyone but someone bent on being contrarian
    for the sake of it, and i will keep referring to it like regardless of
    ur objections

    It also is a valid input that a system defined to answer for "All" needs
    to answer for,

    It seems you world assumes that "all" means maybe most, and "Does Not
    exist" means it might exist.

    That you can answer a question with a non-answer.

    In other words, it is based on LIES.



    Just shows that Turing Machines CAN'T have "self-references" as
    defined in logic.

    It is processing an input that just happens to match a description of
    itself.

    Since that is a legal input, no foul.

    In other words, to try to prohibit this input, means you can't be as
    powerful in computing as a Turing Machine.

    Thus, this "problem" is ESSENTIAL to the nature of the machines, and
    can't be just "fixed".


    Do you think that a compiler can't compile the code for itself?


    Yes, a decider that decides on all inputs (as the decision problem
    requires) needs to be able to decide about the "value" that it
    represents itself.

    In part, the undecidability comes out of the fact that the machinery
    of the system IS powerful enough that we can convert the machines
    into values that can be their inputs.

    Since "Turing Complete" machines can do this, any system that can't
    must be less powerful than "Turing Complete".




    even funnier are all the reactions from chucklefucks resolutely >>>>>>> defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to >>>>>>> kill myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually
    think what it would do).


    algorithm that can determine if a given statement is universally >>>>>>>> valid.
    (Given other results this is equivalent to deciding,
    algorithmically, if
    a given statement is provable.)

    At the time of his now famous paper he was working with Church >>>>>>>> who had
    turned his attention to this as yet unsolved problem.-a At the >>>>>>>> time, most
    mathematicians thought the answer would be "yes".-a Of course the >>>>>>>> first
    step is to capture the notion of an algorithm or process.
    Church came
    up with the lambda calculus, and Turing the abstract machine
    that not
    bears his name.

    However, the idea that mathematics and specifically proofs,
    might be
    automated goes way back to at least Leibniz.

    He worked in defence research

    Not at the time, no.-a He was just a PhD student interested in >>>>>>>> formal
    logic and the intriguing unsolved problems of the time, one of >>>>>>>> which was
    about what can be determined by finite "mechanical" means.

    so I expected he was studying the maximum possible capability >>>>>>>>> of enemy computing infrastructure to help direct research
    resources.












    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From Chris M. Thomasson@chris.m.thomasson.1@gmail.com to comp.theory on Wed Feb 25 14:17:37 2026
    From Newsgroup: comp.theory

    On 2/24/2026 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk>
    writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann posed >>> the question of whether first-order logic was decidable -- was there an

    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all
    twisted up /in the same way/ because of failure to appropriately
    handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.


    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think what
    it would do).

    dart is a low life that likes to tell others to snuff themselves out.
    Laughs about it. Sigh. I suggest plonking that garbage.

    [...]
    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Wed Feb 25 15:30:59 2026
    From Newsgroup: comp.theory

    On 2/25/26 2:17 PM, Chris M. Thomasson wrote:
    On 2/24/2026 6:13 PM, Richard Damon wrote:
    On 2/24/26 7:39 PM, dart200 wrote:
    On 2/24/26 4:09 PM, Ben Bacarisse wrote:
    Tristan Wibberley <tristan.wibberley+netnews2@alumni.manchester.ac.uk> >>>> writes:

    On 20/02/2026 18:11, Ben Bacarisse wrote:
    Turing invented TMs to capture the notion of what is computable

    I'm curious to know how his motives are known.

    At one level there are very clear.-a In 1928, Hilbert and Ackermann
    posed
    the question of whether first-order logic was decidable -- was there an >>>
    and they all gave up on essentially meaninglessly garbage like

    Only meaningless to the stupid.


    und = () -> halts(und) && loop()

    it's kinda funny actually, all these greats getting their panties all
    twisted up /in the same way/ because of failure to appropriately
    handle a lil' self-referential contradiction

    But the "self-reference" isn't actually there in the Turing Formulation.


    even funnier are all the reactions from chucklefucks resolutely
    defending that century long failure as MuH uNQuEsTioNAbLe tRuTH

    actually it's not that funny, or funny at all. makes me want to kill
    myself for being a such a fking EfniEfiA

    Mqybe you should if you are that dumb.

    (Not really, but if you want to keep bringing it up, actually think
    what it would do).

    dart is a low life that likes to tell others to snuff themselves out.
    Laughs about it. Sigh. I suggest plonking that garbage.

    [...]

    twats like u think u deserve respect when u have none to give
    --
    hi, i'm nick! let's end war EfOa

    --- Synchronet 3.21b-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory,alt.messianic,alt.buddha.short.fat.guy on Tue Mar 3 01:44:42 2026
    From Newsgroup: comp.theory

    On 3/3/26 12:55 AM, dart200 wrote:
    On 2/28/26 6:24 PM, Richard Damon wrote:
    On 2/28/26 8:24 PM, dart200 wrote:
    On 2/28/26 2:08 PM, Richard Damon wrote:
    On 2/28/26 12:38 PM, dart200 wrote:
    On 2/28/26 5:21 AM, Richard Damon wrote:
    On 2/27/26 6:09 AM, dart200 wrote:
    On 2/27/26 2:51 AM, Tristan Wibberley wrote:
    On 24/02/2026 21:30, dart200 wrote:
    On 2/24/26 11:38 AM, Tristan Wibberley wrote:
    On 22/02/2026 21:08, dart200 wrote:
    On 2/22/26 12:49 PM, Chris M. Thomasson wrote:
    On 2/22/2026 9:04 AM, dart200 wrote:
    ...
    an effective enumeration of all turing machines was proven on >>>>>>>>>>>>> turing's original paper and can be reused anywhere... >>>>>>>>>>>>
    You think you can test all of them one by one? Don't tell me >>>>>>>>>>>> you think

    yes that's what diagonal proofs do...

    Eh?!

    A test is a procedure! You can't test /all/ of an infinitude >>>>>>>>>> one by one.


    that exactly what turing does in his proof: he defines a
    comptuation
    that enumerates out all the numbers, testing each one of they >>>>>>>>> represent
    a "satisfactory"/"circle-free" machine, and adding that to
    diagonal
    across defined across computable numbers

    it really would be a great exercise to carefully read p247 of >>>>>>>>> turing's
    proof and produce the psuedo-code for the machine H, assuming that >>>>>>>>> machine D exists


    I'll get to it sooner then, because it's mad. Are you sure he >>>>>>>> didn't
    reason quantified over all but phrase it like a procedure for >>>>>>>> what he

    the theory of computation is the theory of such procedures, and >>>>>>> understanding the diagonal procedure is critical to understanding >>>>>>> the *base* contradiction/paradox that the rest of his support for >>>>>>> godel's result is then built on

    And focusing on what is said to be impossible and not changing the >>>>>> problem is also important.

    The problem with the diagonal generation isn't the generation of
    the diagonal itself, but effectively enumerating the enumeration
    in the first place.

    i don't see any indication that turing realized a difference there

    Then you zre just showing your stupidity, because YOU can't tell the
    difference.

    After all, on page 246 he says:

    The computable sequences are therefore not enumerable.

    Here is is SPECIFICALLY talking about the effective enumeration of
    the computable sequences.

    He then points out that he can directly show that the "anti-
    diagonal" of the (non-effectively computed) enumeration can't be
    computed but that "This proof, although perfectly sound, has the
    disadvantage that it may leave the reader with a feeling that 'there
    must be something wrong'".

    it is wrong,

    No, YOU are wrong, as you don't understand what is being done.

    I think he is refering he to the standard anti-diagonal arguement,
    which shows that since for all n, position n differs from the value in
    number n, there can not be any element that matches the anti-diagonal.

    It is just a natural fact of countable infinity, something it seems
    you just don't understand.

    Show how that is actually wrong.

    wow, u know up until now, i thot i fully agreed with turing's short
    diagonal proof, but in writing this post i now find myself in a subtle,
    yet entirely critical disagreement:

    /let an be the n-th computable sequence, and let -an(m) be the m-th
    figure in an. Let +# be the sequence with 1--an(m) as its n-th. figure. Since +# is computable, there exists a number K [== +#] such that 1--an(n)
    = -aK(n) for all n. Putting n = K, we have 1 = 2-aK(K), i.e. 1 is even.
    This is impossible/

    the fallacy here is assuming that because the direct diagonal is
    computable, that one can therefore compute the anti-diagonal using the direct diagonal. the abstract definition makes it look simple, but this ignores the complexities of self-referential analysis (like what turing details on the next page)

    in both methods i have for rectifying the paradox found in the direct diagonal (either (1) filtering TMs or (2) using RTMs), neither can be
    used to then compute the anti-diagonal

    in (1) the algo to compute an inverse diagonal is filtered out like
    turing's paradoxical variation of the direct diagonal would be, and
    there is no analogous non-paradoxical variation that has a hard coded
    value that is inverse to what it does return ... such a concept is
    entirely nonsensical. a function can only return what it does, it can't
    also return the inverse to what it returns eh???

    in (2) the attempt to compute an inverse diagonal with RTMs just fails
    for reasons u'd only understand by working thru the algo urself (p7 of
    re: turing's diagonals)

    the premise:

    /Let +# be the sequence with 1--an(m) as its n-th/

    that sentence there, ben, from p246,

    is the sentence of turing's paper /on computable numbers/ that i start
    to diagree with,

    that sentence just is wrong to imply anything about the computability of
    a number

    /there exists a number K [== +#] such that 1--an(n)
    = -aK(n) for all n/

    and that sentence is *only* /half-true/

    it's *correct* in that can't be computed by a TM,

    whoever a human with a TM could still write it down, but a human could
    never pass that input entirely to a finite running machine eh???

    so i kinda think the ct-thesis is actually cooked in a way

    \_(paa)_/->


    is just not sufficient evidence that such +# is actually computable given the direct diagonal -an()

    one cannot just assume that because the diagonal across computable
    numbers is computable, therefore the anti-diagonal across computable
    numbers is computable...

    He doesn't. You are just showing your stupidity,


    He is proving the enumeration is uncomputable, and without the
    enumeration, you can't compute either of them.


    neither method i have for fixing the diagonal computation across the
    computable numbers can be used to compute the inverse diagonal

    But your method still doesn't let you compute the enumeration, and
    thus you can't actually compute the diagonal.

    Remember, the problem definitions requires that the listing be a
    COMPLETE listing of the computable numbers / machine that compute
    computable numbers, in some definite order.

    If your enumeration isn't complete, your diagonal isn't correct.


    so while i agree with turing that the anti-diagonal is not
    computable, i don't agree that the normal diagonal is not computable

    Why?

    How does D decide on the original H?

    Your modified H still needs a correct D to decide on all the other
    machines, including his original H that doesn't use your "trick"




    But instead, he can prove with a more obvious process, that the
    Decider "D" that could be used to effectively enumerate the sequence
    of machine that produce computable numbers can not esit.

    Thus, he clearly knows the difference, but is pointing out that the
    attempt to compute the diagonal clearly reveals the issue with
    effectively enumerating the sequences.

    well, he didn't consider that perhaps the proper algo for computing
    the diagonal can avoid the paradox on itself ...

    But it doesn't.

    Your just don't understand that D just can't correctly decide on his
    given H.

    no idea why ur claiming that

    i clearly understand that D cannot decide correctly on turing's H,
    because my response to this is that D does not need to decide correctly
    on H to compute a diagonal


    It doesn't matter that your new H doesn't get stuck on itself, it will
    still error on Turing's H.

    turing's H, as it stands, doesn't even exist my dude. he doesn't specify what D (or H) needs to do when encountering the /undecidable input/ of
    H, so therefore both D and H are an incomplete specifications of a machine


    IF D is wrong by deciding it is not circle free, then your H will
    compute the wrong diagonal, as the resulting version of his H WILL be
    circle free (since it never tries to simulate itself) and thus DOES
    produce an computable number that your computation misses.

    Or, if that D is wrong by decing it IS circle free, then when you H
    tries to process it, it will get stuck in the infinite loop.

    The problem is that in stepping through the machines in order, you
    WILL hit these actual machines built on your erroneous D (your D must
    have this flaw, as no D without exists), and thus you will be wrong on
    THAT input. IT doesn't matter if you get a good answer for yourself.


    idk what he would have said about it, but prolly something more
    substantial than just calling me ignorant repeatedly

    I doubt it.

    He likely would have gotten frustrated by your idiodic assertion of
    bad logic. You would have likely been escorted out of the meeting as
    showing you were unqualified and being a distraction.



    Something that seems to be beyond your ignorant understanding.



    H shows that *IF* you can make that enumeration, you can make the >>>>>> diagonal, and thus the anti-diagonal. The problem is you can't
    make that enumeration, and assuming you can just shows unsoundness. >>>>>
    interestingly: one can only fix the direct diagonal computation
    like this

    u can't do an analogous fix for the inverse/anti-diagonal
    computation. it's not possible hard code a machine to return an
    inverted value, a machine can only return what it does, not the
    inverse of what it does...

    so if we can filter out paradoxes from the enumeration, that will
    leave a direct diagonal computation extant in that filtered (yet
    still turing complete list), while any attempt to compute an
    inverse diagonal will not be

    But the problem is that "paradoxical machines" don't exist in
    isolation, but only in relationship to a given machine trying to
    decide them.

    right. so if ur constructing a diagonal across computable numbers
    then u only need to filter out paradoxes in regards to the classifier
    that classifies them as a "satisfactory" number

    Right, which he shows can not be done.

    please do quote where turing shows we can't filter out such paradoxes...

    (also why do always just make random assertions???)



    any machine which *is not* "satisfactory" OR *is not* classifiable as
    satisfactory by said classifier... can just be skipped

    No, it can only skip those that are not satisfactory, not those that
    are but it can not classify as such, or your enumeration will not be
    complete, and thus just in error.

    Thus, it needs to be able to correctly classify ALL machines (as it
    will be asked about all machines as it counts through all the
    descriptions) and thus Turing's H *WILL* be asked about.


    similarly if u want to go a step further an filter out computable
    numbers already included on this diagonal, any machine which either
    *is* computably equivalent OR *is not* classifiable in regards to
    *any* machine already the list... can just be skipped

    Nope, you can't skip some machines, as you then might lose some of the
    computable numbers.


    see you can't compute a diagonal across *all* /machines/, with said
    machines, but u can compute a diagonal across *all* /computable numbers/

    Nope,

    Since the enumeration of ALL Computable numbers can't be done, since
    ALL classifiers that attempt it will make an error, you can't do what
    you want to do.

    nah, (a) computing an enumeration of all /computable numbers/ is not the same thing as (b) computing the enumeration of all machines that compute computable numbers. (b) necessarily has duplicates while (a) does not
    need them. turing's paper wrongly conflates (a) with (b)

    i'm pretty sure (a) can be done with TMs

    (b) probably can't be done with TMs



    yes, i still do need to prove my thesis that for any paradoxical
    machine, there exists a functionally equivalent machine without such
    paradox

    And the problem is that your "paradoxical" isn't actually a definable
    property (let alone computable). Part of the problem is that if you
    look at just a machine description, it doesn't (necessarily) tell you
    about the use of an "interface" as that use of an interface can be
    just inlined, leaving nothing "in the code" to show it exists.

    i'm sorry, are you actually saying the machine description does not
    describe what the machine does???

    lol





    His specified H, with an actually (incorrect) implementation of D
    (which is all that CAN exist) will either be circle-free and thus
    generate a number (but its D said it isn't, and thus omitted a valid
    machine from the list) or it isn't circle-free, and fails to computa
    a number, and thus should have been omitted from the list but wasn't.

    Thus any H that ACTUALLY EXISTS, isn't a "paradox", it is just built
    on an assuption in error.


    so despite turing's worries, the existence of a diagonal
    computation does not actually then imply the existence of an anti-
    diagonal computation, due the same particular self-referential
    weirdness that stumped turing the first place

    But there is no actuall SELF-REFERENCE, so your logic is just based
    on ERROR.

    Your attempt to REDEFINE self-reference to mean processing a number
    that happens to represent yourself means that you you system "ALL"
    doesn't actually mean ALL, and thus is just ill-defined.

    i'm sorry, you have an issue with me labeling a number that directly
    refers to yourself, as a "self-reference" ???

    Sure, because it is just a number. The problem is that you still have

    it's a *specific* number that has the currently running machine encoded
    into it, it's not "just" a number whatever that means

    problems with all the "equivalent" machines that have different numbers.

    those are references to functionally equivalent machines, not self- references


    It may let you filter out the simplest case used in the proofs, but
    doesn't solve the actual problem, as the "Machine Number" doesn't
    actually fully identify the problematic cases.

    that's not actually true. you can't meaningfully paradox the paradox detector while filter *out* paradoxes to point of creating an actual
    problem

    let halts be a halting classifier that takes an input machines and
    returns true/false whether it halts or not

    let paradox_free be a paradox filtering classifier that takes an input classifier, and an input machine to search for paradoxes in regards to: paradox_free(classifier, machine)

    for example if take our basic halting paradox:

    -a und = () -> halts(und) loop()

    then:

    -a paradox_free(halts, und) -> TRUE

    but you might think one still can bamboozle the paradox classifier like such:

    -a undp = () -> {
    -a-a-a if ( paradox_free(halts, undp) == TRUE )
    -a-a-a-a-a if ( halts(undp) ) loop()
    -a }

    here if paradox_free(halts,undp) -> FALSE, then the if case run a
    halting paradox form making it /undecidable input/ to halts().

    however, if paradox_free(halts,undp) -> TRUE, then that if case doesn't
    run and clearly halts(undp) -> TRUE without issues

    BUT THAT'S FINE FOR OUR USE CASE, no paradox found within a machine is necessary for the particular input->output computation being done, so utilizing the return paradox_free(halts,undp) -> FALSE to filter out paradox_free() paradoxes as well as halts() paradoxes will suffice to produce a turing complete subset of machines that can be totally
    classified by halts()

    i think a major problem with the consensus perspective is an expectation
    of a certain cleanliness in the logic as being the only possibility here
    to compute what we want. TM's can't achieve that, but something slightly messier can.




    dear future: u see this shit i'm trying work with???? sorry i didn't
    manage to make progress any faster, but i'm literally dragging
    toddlers kicking and screaming at this point jeez...


    In other words, in your world "ALL" isn't a word that can actually be
    used.

    You don't understand that you need to use actual sound logic,

    lol, i can only EfOA someone would engage in sound logic with me





    ...which u won't find interesting, but i can't fix a lack of curiosity >>>>>
    Efn+



    expected was an inexpert audience, supposing that experts will >>>>>>>> recognise
    the relevant mapping to universal quantification?





    This just shows that you real problem is you don't understand what
    the actual problem is, and your world is just build on things that
    are lies.

    IT seems that fundamentally, your world doesn't actually have
    "computations" as you don't understand the basic requirement that
    they need to be fully defined in the actions they do.



    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the little crank that could
    --- Synchronet 3.21d-Linux NewsLink 1.2