• ct thesis is prolly fucked

    From dart200@user7160@newsgrouper.org.invalid to comp.theory on Wed Jul 15 21:34:41 2026
    From Newsgroup: comp.theory

    upon picking apart the abject nonsense that is recursive undecidability,
    so that i can discuss what undecidability _actually_ looks like within computing,

    i find myself stumbling into a proof that turing machines as a model are
    not capable of computing everything that can be computed mechanically

    i guess we'll see where the paper ultimately leads for sure, but things
    are cooking rLiN+A
    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the lil crank that could

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Dude@user2891@newsgrouper.org.invalid to comp.theory on Wed Aug 5 21:29:14 2026
    From Newsgroup: comp.theory


    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive undecidability,
    so that i can discuss what undecidability _actually_ looks like within computing,

    i find myself stumbling into a proof that turing machines as a model are
    not capable of computing everything that can be computed mechanically

    i guess we'll see where the paper ultimately leads for sure, but things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating from your kitchen table.

    As an IT professional, I can say without the least hesitation, there's not many machines that can compare with a Cray.

    Cray Computers are optimized for massive mathematical arrays rather than standard scalar math. Crays are heavily utilized in nuclear research,
    aerospace design, codebreaking, and global satellite weather prediction.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to comp.theory on Sat Aug 8 01:18:01 2026
    From Newsgroup: comp.theory

    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive undecidability,
    so that i can discuss what undecidability _actually_ looks like within
    computing,

    i find myself stumbling into a proof that turing machines as a model are
    not capable of computing everything that can be computed mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a Turing machine, as a subject for your thesis? I believe that's been done, so
    the question is, what exactly is it about mechanical computing that
    cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating from your kitchen table.

    That is a matter of perspective. The fruit laptop on the kitchen table
    can be thought of as a terminal, or even just a typewriter, for the
    Cray.


    As an IT professional, I can say without the least hesitation, there's not many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather than standard scalar math. Crays are heavily utilized in nuclear research, aerospace design, codebreaking, and global satellite weather prediction.


    So if I understand Finseth's book about how to code an Emacs correctly,
    a Cray supercomputer would be massively hampered were you to implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to comp.theory on Wed Aug 12 19:34:52 2026
    From Newsgroup: comp.theory

    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive undecidability, >>> so that i can discuss what undecidability _actually_ looks like within
    computing,

    i find myself stumbling into a proof that turing machines as a model are >>> not capable of computing everything that can be computed mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a Turing machine, as a subject for your thesis? I believe that's been done, so
    the question is, what exactly is it about mechanical computing that
    cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating from
    your
    kitchen table.

    That is a matter of perspective. The fruit laptop on the kitchen table
    can be thought of as a terminal, or even just a typewriter, for the
    Cray.


    As an IT professional, I can say without the least hesitation, there's
    not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather than
    standard scalar math. Crays are heavily utilized in nuclear research,
    aerospace design, codebreaking, and global satellite weather prediction.


    So if I understand Finseth's book about how to code an Emacs correctly,
    a Cray supercomputer would be massively hampered were you to implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Wed Aug 12 22:15:15 2026
    From Newsgroup: comp.theory

    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like within >>>> computing,

    i find myself stumbling into a proof that turing machines as a model
    are
    not capable of computing everything that can be computed mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a Turing
    machine, as a subject for your thesis?-a I believe that's been done, so
    the question is, what exactly is it about mechanical computing that
    cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but things >>>> are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating from
    your
    kitchen table.

    That is a matter of perspective.-a The fruit laptop on the kitchen table
    can be thought of as a terminal, or even just a typewriter, for the
    Cray.


    As an IT professional, I can say without the least hesitation, there's
    not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather than >>> standard scalar math. Crays are heavily utilized in nuclear research,
    aerospace design, codebreaking, and global satellite weather prediction.


    So if I understand Finseth's book about how to code an Emacs correctly,
    a Cray supercomputer would be massively hampered were you to implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit when it
    comes infinite ...

    turing for example tried to extend logical systems into transfinite
    ordinals in an attempt to get around incompleteness. no idea what kind
    of axiom would realistically come past an infinite amount of them, but
    this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing, which
    in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy
    --
    why are we god?
    let's end war EfOa

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to comp.theory on Thu Aug 13 10:12:04 2026
    From Newsgroup: comp.theory

    On 08/12/2026 10:15 PM, dart200 wrote:
    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like within >>>>> computing,

    i find myself stumbling into a proof that turing machines as a
    model are
    not capable of computing everything that can be computed mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a Turing >>> machine, as a subject for your thesis? I believe that's been done, so
    the question is, what exactly is it about mechanical computing that
    cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but
    things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating from >>>> your
    kitchen table.

    That is a matter of perspective. The fruit laptop on the kitchen table
    can be thought of as a terminal, or even just a typewriter, for the
    Cray.


    As an IT professional, I can say without the least hesitation, there's >>>> not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather
    than
    standard scalar math. Crays are heavily utilized in nuclear research,
    aerospace design, codebreaking, and global satellite weather
    prediction.


    So if I understand Finseth's book about how to code an Emacs correctly,
    a Cray supercomputer would be massively hampered were you to implement a >>> text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit when it
    comes infinite ...

    turing for example tried to extend logical systems into transfinite
    ordinals in an attempt to get around incompleteness. no idea what kind
    of axiom would realistically come past an infinite amount of them, but
    this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing, which
    in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy


    Why would you do that?

    It's kind of like driving a car, and knowing the limits,
    since one never knows the limits, yet is always finding the limits, incompleteness is simply beyond a limit, and not being cognizant
    of it, if not quite sure where it is, risks hitting its wall,
    or driving over its cliff.

    It's like the very notion of analysis, making
    restriction-of-comprehension keeps some things simple, other things un-available to reason. Somebody else needn't adopt that restriction,
    and then it's like a rooster locking itself in the coop.

    Mirimanoff points out that a model of finite ordinals is extra-ordinary, Russell wishes it away, the "Russell-ian retro-thesis",
    it's natural that infinity is "in", and that models of ordinals
    or integers start as both bounded-fragments and
    extraordinary-extensions, the "standard" model not even existing except as
    a limit of those.

    The classical expositions of the super-classical, usually considered
    to start with Zeno's account of infinite-divisbility and summability
    and motion in time, and the geometric series, gives reasons why that
    for infinitary-analysis, that's about the only place to start that
    isn't wacky, and it makes of itself accounts of why induction fails
    where deduction succeeds, and about complementary duals, and the
    great account of structure and geometry and arithmetic, and continuity
    and infinity, up-front.

    Then, that there are more models of laws of large numbers, and
    that naive induction is basically finitistic and only knows a
    law of small numbers, or "weak" induction, compared to the "strong"
    induction which is given an account by the existence of a space
    and its structure, that the "a priori" of "strong induction" is
    due classical-expositions of super-classical results after the
    double-reductio and the ad-infinitum instead of the ad-absurdam,
    about the "ab-absurdam", that numbers and forms always have those.


    Then, retro-finitism or ultra-finitism is naive, and incomplete.



    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Thu Aug 13 18:29:14 2026
    From Newsgroup: comp.theory

    On 8/13/26 10:12 AM, Ross Finlayson wrote:
    On 08/12/2026 10:15 PM, dart200 wrote:
    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like
    within
    computing,

    i find myself stumbling into a proof that turing machines as a
    model are
    not capable of computing everything that can be computed mechanically >>>>

    Have you considered how to model the /Antikythera Mechanism/ on a
    Turing
    machine, as a subject for your thesis?-a I believe that's been done, so >>>> the question is, what exactly is it about mechanical computing that
    cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but
    things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating from >>>>> your
    kitchen table.

    That is a matter of perspective.-a The fruit laptop on the kitchen table >>>> can be thought of as a terminal, or even just a typewriter, for the
    Cray.


    As an IT professional, I can say without the least hesitation, there's >>>>> not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't comment on >>>> that.


    Cray Computers are optimized for massive mathematical arrays rather
    than
    standard scalar math. Crays are heavily utilized in nuclear research, >>>>> aerospace design, codebreaking, and global satellite weather
    prediction.


    So if I understand Finseth's book about how to code an Emacs correctly, >>>> a Cray supercomputer would be massively hampered were you to
    implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit when it
    comes infinite ...

    turing for example tried to extend logical systems into transfinite
    ordinals in an attempt to get around incompleteness. no idea what kind
    of axiom would realistically come past an infinite amount of them, but
    this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing, which
    in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy


    Why would you do that?

    because erroneously proving a limit that does not actually exist will artificially limit our potential to wield computing as an applied technique

    like specifically when it comes to our practical application of
    computing, we do not prove what our computations do. we might test
    various inputs/outputs combos, but this is quite a bit inferior to
    proving semantics across their entire input possibility space (and not
    just brute forcing that proof)


    It's kind of like driving a car, and knowing the limits,
    since one never knows the limits, yet is always finding the limits, incompleteness is simply beyond a limit, and not being cognizant
    of it, if not quite sure where it is, risks hitting its wall,
    or driving over its cliff.

    i'm not sure how framing currently intractable problems as actually
    tractable (without limiting the domain!) risks driving over some kind of existential cliff in this case

    undecidability within computing is entirely founded in the problems of computing the semantic properties of computations, unlocking those kinds
    of problems as actually tractable seems like possibly a huge boon to the advancement of computing


    It's like the very notion of analysis, making
    restriction-of-comprehension keeps some things simple, other things un-available to reason. Somebody else needn't adopt that restriction,
    and then it's like a rooster locking itself in the coop.

    i'm not proposing a reduction in power to computing, if anything this
    would be an expansion in descriptive power


    Mirimanoff points out that a model of finite ordinals is extra-ordinary, Russell wishes it away, the "Russell-ian retro-thesis",
    it's natural that infinity is "in", and that models of ordinals
    or integers start as both bounded-fragments and
    extraordinary-extensions, the "standard" model not even existing except as
    a limit of those.

    The classical expositions of the super-classical, usually considered
    to start with Zeno's account of infinite-divisbility and summability
    and motion in time, and the geometric series, gives reasons why that
    for infinitary-analysis, that's about the only place to start that
    isn't wacky, and it makes of itself accounts of why induction fails
    where deduction succeeds, and about complementary duals, and the
    great account of structure and geometry and arithmetic, and continuity
    and infinity, up-front.

    i have a hard time parsing what u mean even with gemenigpts help, but
    please do remember the /theory of computing/ is inherently limited to
    the cardinality of natural numbers, as mechanical computations must be
    mapped to the finite machine description which compute them


    Then, that there are more models of laws of large numbers, and
    that naive induction is basically finitistic and only knows a
    law of small numbers, or "weak" induction, compared to the "strong"
    induction which is given an account by the existence of a space
    and its structure, that the "a priori" of "strong induction" is
    due classical-expositions of super-classical results after the double-reductio and the ad-infinitum instead of the ad-absurdam,
    about the "ab-absurdam", that numbers and forms always have those.


    Then, retro-finitism or ultra-finitism is naive, and incomplete.
    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the lil crank that could
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to comp.theory on Fri Aug 14 07:27:14 2026
    From Newsgroup: comp.theory

    On 08/13/2026 06:29 PM, dart200 wrote:
    On 8/13/26 10:12 AM, Ross Finlayson wrote:
    On 08/12/2026 10:15 PM, dart200 wrote:
    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like
    within
    computing,

    i find myself stumbling into a proof that turing machines as a
    model are
    not capable of computing everything that can be computed
    mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a
    Turing
    machine, as a subject for your thesis? I believe that's been done, so >>>>> the question is, what exactly is it about mechanical computing that
    cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but >>>>>>> things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating
    from
    your
    kitchen table.

    That is a matter of perspective. The fruit laptop on the kitchen
    table
    can be thought of as a terminal, or even just a typewriter, for the
    Cray.


    As an IT professional, I can say without the least hesitation,
    there's
    not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't
    comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather >>>>>> than
    standard scalar math. Crays are heavily utilized in nuclear research, >>>>>> aerospace design, codebreaking, and global satellite weather
    prediction.


    So if I understand Finseth's book about how to code an Emacs
    correctly,
    a Cray supercomputer would be massively hampered were you to
    implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit when it
    comes infinite ...

    turing for example tried to extend logical systems into transfinite
    ordinals in an attempt to get around incompleteness. no idea what kind
    of axiom would realistically come past an infinite amount of them, but
    this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing, which
    in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy


    Why would you do that?

    because erroneously proving a limit that does not actually exist will artificially limit our potential to wield computing as an applied technique

    like specifically when it comes to our practical application of
    computing, we do not prove what our computations do. we might test
    various inputs/outputs combos, but this is quite a bit inferior to
    proving semantics across their entire input possibility space (and not
    just brute forcing that proof)


    It's kind of like driving a car, and knowing the limits,
    since one never knows the limits, yet is always finding the limits,
    incompleteness is simply beyond a limit, and not being cognizant
    of it, if not quite sure where it is, risks hitting its wall,
    or driving over its cliff.

    i'm not sure how framing currently intractable problems as actually
    tractable (without limiting the domain!) risks driving over some kind of existential cliff in this case

    undecidability within computing is entirely founded in the problems of computing the semantic properties of computations, unlocking those kinds
    of problems as actually tractable seems like possibly a huge boon to the advancement of computing


    It's like the very notion of analysis, making
    restriction-of-comprehension keeps some things simple, other things
    un-available to reason. Somebody else needn't adopt that restriction,
    and then it's like a rooster locking itself in the coop.

    i'm not proposing a reduction in power to computing, if anything this
    would be an expansion in descriptive power


    Mirimanoff points out that a model of finite ordinals is extra-ordinary,
    Russell wishes it away, the "Russell-ian retro-thesis",
    it's natural that infinity is "in", and that models of ordinals
    or integers start as both bounded-fragments and
    extraordinary-extensions, the "standard" model not even existing
    except as
    a limit of those.

    The classical expositions of the super-classical, usually considered
    to start with Zeno's account of infinite-divisbility and summability
    and motion in time, and the geometric series, gives reasons why that
    for infinitary-analysis, that's about the only place to start that
    isn't wacky, and it makes of itself accounts of why induction fails
    where deduction succeeds, and about complementary duals, and the
    great account of structure and geometry and arithmetic, and continuity
    and infinity, up-front.

    i have a hard time parsing what u mean even with gemenigpts help, but
    please do remember the /theory of computing/ is inherently limited to
    the cardinality of natural numbers, as mechanical computations must be
    mapped to the finite machine description which compute them


    Then, that there are more models of laws of large numbers, and
    that naive induction is basically finitistic and only knows a
    law of small numbers, or "weak" induction, compared to the "strong"
    induction which is given an account by the existence of a space
    and its structure, that the "a priori" of "strong induction" is
    due classical-expositions of super-classical results after the
    double-reductio and the ad-infinitum instead of the ad-absurdam,
    about the "ab-absurdam", that numbers and forms always have those.


    Then, retro-finitism or ultra-finitism is naive, and incomplete.


    Hm. Thanks for writing.

    Agreeably, accounts of the _unbounded_, of the _finite_, keep
    things "sensible, fungible, and tractable", and that accounts
    of _completions_, in the _infinite_, are due descriptions by
    super-classical results like what give the geometric series,
    or Zeno's arguments either way _both existing_ when induction
    either way makes a counter-development/counter-example to the other,
    this is about "weak and strong induction", that induction and infinite induction it's usual given account of the base case then inductive case,
    is "weak", since it's _completion_, doesn't have an "actual
    infinite".


    For the theory of computation, a usual idea is that there are
    "Turing machines" and then "Zeno machines", then, that like
    there are the "digital" and "analog", the "discrete" and "continuous",
    that it's independent number theory which of large, larger, and largest
    laws of large numbers apply, for inductive, infinite, and continuum limits.

    There are at least three kinds of limits, then that if infinite limits
    and continuum limits require the "strong induction", i.e., weak
    induction plus also another reason why the completion occurs,
    then makes for "completeness" and "measure" after "density".

    The unbounded readily gives density, it's agreeable.


    I suppose it's been called "non-standard", like "non-standard models
    of integers" or "non-standard models of probability", yet, that's
    because the entire linear curriculum mostly doesn't have an account
    of deductive analysis at all, since, trivially enough, it's provides
    a direct and immediate counter-development and counter-example to
    any plain course of the naive (or weak) induction at all, which has
    been known since antiquity about why the ancients said truth was
    "discovered" instead of "invented".

    Approximation algorithms abound, to be sure, and approximations
    always have a nominally non-zero error term, and modeling the error term
    or modeling the error bounds of approximations, include for
    where the error term gets away from the approximation itself.


    Then, the "as-if" or "almost", like the "almost-all",
    "almost-everywhere", "almost-periodic", "almost-analytic", the sorts
    of "almost-approximate", that's naturally enough more of a
    "what-if" than an "as-if".


    So, "modeling the error bounds" is usually what's involved in
    analysis since here the study of things is called "dynamical
    modeling" not "chaos theory", since math is not indeterministic,
    then about singularity theory that singularities in a singularity
    theory are branches in a multiplicity theory, just taking the extra-book-keeping when there's enough scratch-space to write
    out the numerical method or approximative algorithm, or making
    an account of how mathematics does that in actual numerical resources,
    or for the usual accounts of incompleteness (or, inconsistency if
    you'd rather not have incompleteness) of the ordinary, and somehow
    the completeness and the consistency in the extra-ordinary,
    of the infinite and continuous.

    So, Church-Turing thesis is prolly not fritzed, though that
    it is though merely a "fragment" or the "ordinary", and,
    in the extra-ordinary then thusly, it's _independent_, the ordinary.
    (The "standard" in one sense of the ordinary is "non-standard"
    the other, like "standard infinitesimals" or "standard integers",
    one giving a clock-arithmetic the other a field-arithmetic,
    and only agreeing about the measure of [0,1].)





    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From dart200@user7160@newsgrouper.org.invalid to comp.theory on Sat Aug 15 14:38:02 2026
    From Newsgroup: comp.theory

    On 8/14/26 7:27 AM, Ross Finlayson wrote:
    On 08/13/2026 06:29 PM, dart200 wrote:
    On 8/13/26 10:12 AM, Ross Finlayson wrote:
    On 08/12/2026 10:15 PM, dart200 wrote:
    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like >>>>>>>> within
    computing,

    i find myself stumbling into a proof that turing machines as a >>>>>>>> model are
    not capable of computing everything that can be computed
    mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a
    Turing
    machine, as a subject for your thesis?-a I believe that's been
    done, so
    the question is, what exactly is it about mechanical computing that >>>>>> cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but >>>>>>>> things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating >>>>>>> from
    your
    kitchen table.

    That is a matter of perspective.-a The fruit laptop on the kitchen >>>>>> table
    can be thought of as a terminal, or even just a typewriter, for the >>>>>> Cray.


    As an IT professional, I can say without the least hesitation,
    there's
    not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't
    comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather >>>>>>> than
    standard scalar math. Crays are heavily utilized in nuclear
    research,
    aerospace design, codebreaking, and global satellite weather
    prediction.


    So if I understand Finseth's book about how to code an Emacs
    correctly,
    a Cray supercomputer would be massively hampered were you to
    implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit when it >>>> comes infinite ...

    turing for example tried to extend logical systems into transfinite
    ordinals in an attempt to get around incompleteness. no idea what kind >>>> of axiom would realistically come past an infinite amount of them, but >>>> this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing, which >>>> in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy


    Why would you do that?

    because erroneously proving a limit that does not actually exist will
    artificially limit our potential to wield computing as an applied
    technique

    like specifically when it comes to our practical application of
    computing, we do not prove what our computations do. we might test
    various inputs/outputs combos, but this is quite a bit inferior to
    proving semantics across their entire input possibility space (and not
    just brute forcing that proof)


    It's kind of like driving a car, and knowing the limits,
    since one never knows the limits, yet is always finding the limits,
    incompleteness is simply beyond a limit, and not being cognizant
    of it, if not quite sure where it is, risks hitting its wall,
    or driving over its cliff.

    i'm not sure how framing currently intractable problems as actually
    tractable (without limiting the domain!) risks driving over some kind of
    existential cliff in this case

    undecidability within computing is entirely founded in the problems of
    computing the semantic properties of computations, unlocking those kinds
    of problems as actually tractable seems like possibly a huge boon to the
    advancement of computing


    It's like the very notion of analysis, making
    restriction-of-comprehension keeps some things simple, other things
    un-available to reason. Somebody else needn't adopt that restriction,
    and then it's like a rooster locking itself in the coop.

    i'm not proposing a reduction in power to computing, if anything this
    would be an expansion in descriptive power


    Mirimanoff points out that a model of finite ordinals is extra-ordinary, >>> Russell wishes it away, the "Russell-ian retro-thesis",
    it's natural that infinity is "in", and that models of ordinals
    or integers start as both bounded-fragments and
    extraordinary-extensions, the "standard" model not even existing
    except as
    a limit of those.

    The classical expositions of the super-classical, usually considered
    to start with Zeno's account of infinite-divisbility and summability
    and motion in time, and the geometric series, gives reasons why that
    for infinitary-analysis, that's about the only place to start that
    isn't wacky, and it makes of itself accounts of why induction fails
    where deduction succeeds, and about complementary duals, and the
    great account of structure and geometry and arithmetic, and continuity
    and infinity, up-front.

    i have a hard time parsing what u mean even with gemenigpts help, but
    please do remember the /theory of computing/ is inherently limited to
    the cardinality of natural numbers, as mechanical computations must be
    mapped to the finite machine description which compute them


    Then, that there are more models of laws of large numbers, and
    that naive induction is basically finitistic and only knows a
    law of small numbers, or "weak" induction, compared to the "strong"
    induction which is given an account by the existence of a space
    and its structure, that the "a priori" of "strong induction" is
    due classical-expositions of super-classical results after the
    double-reductio and the ad-infinitum instead of the ad-absurdam,
    about the "ab-absurdam", that numbers and forms always have those.


    Then, retro-finitism or ultra-finitism is naive, and incomplete.


    Hm. Thanks for writing.

    Agreeably, accounts of the _unbounded_, of the _finite_, keep
    things "sensible, fungible, and tractable", and that accounts
    of _completions_, in the _infinite_, are due descriptions by
    super-classical results like what give the geometric series,
    or Zeno's arguments either way _both existing_ when induction
    either way makes a counter-development/counter-example to the other,
    this is about "weak and strong induction", that induction and infinite induction it's usual given account of the base case then inductive case,
    is "weak", since it's _completion_, doesn't have an "actual
    infinite".


    For the theory of computation, a usual idea is that there are
    "Turing machines" and then "Zeno machines", then, that like
    there are the "digital" and "analog", the "discrete" and "continuous",
    that it's independent number theory which of large, larger, and largest
    laws of large numbers apply, for inductive, infinite, and continuum limits.

    There are at least three kinds of limits, then that if infinite limits
    and continuum limits require the "strong induction", i.e., weak
    induction plus also another reason why the completion occurs,
    then makes for "completeness" and "measure" after "density".

    The unbounded readily gives density, it's agreeable.


    I suppose it's been called "non-standard", like "non-standard models
    of integers" or "non-standard models of probability", yet, that's
    because the entire linear curriculum mostly doesn't have an account
    of deductive analysis at all, since, trivially enough, it's provides
    a direct and immediate counter-development and counter-example to
    any plain course of the naive (or weak) induction at all, which has
    been known since antiquity about why the ancients said truth was
    "discovered" instead of "invented".

    Approximation algorithms abound, to be sure, and approximations
    always have a nominally non-zero error term, and modeling the error term
    or modeling the error bounds of approximations, include for
    where the error term gets away from the approximation itself.


    Then, the "as-if" or "almost", like the "almost-all",
    "almost-everywhere", "almost-periodic", "almost-analytic", the sorts
    of "almost-approximate", that's naturally enough more of a
    "what-if" than an "as-if".


    So, "modeling the error bounds" is usually what's involved in
    analysis since here the study of things is called "dynamical
    modeling" not "chaos theory", since math is not indeterministic,
    then about singularity theory that singularities in a singularity
    theory are branches in a multiplicity theory, just taking the extra-book-keeping when there's enough scratch-space to write
    out the numerical method or approximative algorithm, or making
    an account of how mathematics does that in actual numerical resources,
    or for the usual accounts of incompleteness (or, inconsistency if
    you'd rather not have incompleteness) of the ordinary, and somehow
    the completeness and the consistency in the extra-ordinary,
    of the infinite and continuous.

    So, Church-Turing thesis is prolly not fritzed, though that
    it is though merely a "fragment" or the "ordinary", and,

    ross, i'm not using a model of hypercomputation to usurp the ct thesis.
    as we don't actually know how to mechanically implement those, they
    cannot be used by a human to compute something a turing machine cannot,
    and therefore do not refute it.

    recall that the ct thesis is thus:

    | A function is effectively calculable by a human being
    | _iff_ it can be computed by a turing machine.

    the ct thesis claims actual mechanical computation that a human can do
    is limited to what can be expressed within the turing machine model.
    refuting the ct thesis involves demonstrating a mechanical process that utilizes an idealized human agent to compute something which a turing
    machine simply cannot express.

    the refutation is not algorithmically novel or hard, it's more due to
    the fact that the human operation exists outside what can be directly referenced by the turing machine model. yes, we can simulate the human operation within the turing machine model, and that simulation will be
    limited by what turing machines can express, but a human agent
    mechanically doing that same thing just is not. i'm sure u won't be
    convinced by my paragraphs here, the full argument will be in a paper
    i'm finishing up.

    the refutation is again, not algorithmically hard, it's moreso tied to
    the same self-referential paradox which stumped turing into establishing limits to turing machine computation in the first place, back on the
    1936 paper /on computable numbers/

    in the extra-ordinary then thusly, it's _independent_, the ordinary.
    (The "standard" in one sense of the ordinary is "non-standard"
    the other, like "standard infinitesimals" or "standard integers",
    one giving a clock-arithmetic the other a field-arithmetic,
    and only agreeing about the measure of [0,1].)





    --
    arising us out of the computing dark ages,
    please excuse my pseudo-pyscript,
    ~ the lil crank that could
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to comp.theory on Mon Aug 17 06:23:03 2026
    From Newsgroup: comp.theory

    On 16/08/2026 5:38 AM, dart200 wrote:
    On 8/14/26 7:27 AM, Ross Finlayson wrote:
    On 08/13/2026 06:29 PM, dart200 wrote:
    On 8/13/26 10:12 AM, Ross Finlayson wrote:
    On 08/12/2026 10:15 PM, dart200 wrote:
    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like >>>>>>>>> within
    computing,

    i find myself stumbling into a proof that turing machines as a >>>>>>>>> model are
    not capable of computing everything that can be computed
    mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a >>>>>>> Turing
    machine, as a subject for your thesis?-a I believe that's been
    done, so
    the question is, what exactly is it about mechanical computing that >>>>>>> cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but >>>>>>>>> things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating >>>>>>>> from
    your
    kitchen table.

    That is a matter of perspective.-a The fruit laptop on the kitchen >>>>>>> table
    can be thought of as a terminal, or even just a typewriter, for the >>>>>>> Cray.


    As an IT professional, I can say without the least hesitation, >>>>>>>> there's
    not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't
    comment on
    that.


    Cray Computers are optimized for massive mathematical arrays rather >>>>>>>> than
    standard scalar math. Crays are heavily utilized in nuclear
    research,
    aerospace design, codebreaking, and global satellite weather
    prediction.


    So if I understand Finseth's book about how to code an Emacs
    correctly,
    a Cray supercomputer would be massively hampered were you to
    implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply
    doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit when it >>>>> comes infinite ...

    turing for example tried to extend logical systems into transfinite
    ordinals in an attempt to get around incompleteness. no idea what kind >>>>> of axiom would realistically come past an infinite amount of them, but >>>>> this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing,
    which
    in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy


    Why would you do that?

    because erroneously proving a limit that does not actually exist will
    artificially limit our potential to wield computing as an applied
    technique

    like specifically when it comes to our practical application of
    computing, we do not prove what our computations do. we might test
    various inputs/outputs combos, but this is quite a bit inferior to
    proving semantics across their entire input possibility space (and not
    just brute forcing that proof)


    It's kind of like driving a car, and knowing the limits,
    since one never knows the limits, yet is always finding the limits,
    incompleteness is simply beyond a limit, and not being cognizant
    of it, if not quite sure where it is, risks hitting its wall,
    or driving over its cliff.

    i'm not sure how framing currently intractable problems as actually
    tractable (without limiting the domain!) risks driving over some kind of >>> existential cliff in this case

    undecidability within computing is entirely founded in the problems of
    computing the semantic properties of computations, unlocking those kinds >>> of problems as actually tractable seems like possibly a huge boon to the >>> advancement of computing


    It's like the very notion of analysis, making
    restriction-of-comprehension keeps some things simple, other things
    un-available to reason. Somebody else needn't adopt that restriction,
    and then it's like a rooster locking itself in the coop.

    i'm not proposing a reduction in power to computing, if anything this
    would be an expansion in descriptive power


    Mirimanoff points out that a model of finite ordinals is extra-
    ordinary,
    Russell wishes it away, the "Russell-ian retro-thesis",
    it's natural that infinity is "in", and that models of ordinals
    or integers start as both bounded-fragments and
    extraordinary-extensions, the "standard" model not even existing
    except as
    a limit of those.

    The classical expositions of the super-classical, usually considered
    to start with Zeno's account of infinite-divisbility and summability
    and motion in time, and the geometric series, gives reasons why that
    for infinitary-analysis, that's about the only place to start that
    isn't wacky, and it makes of itself accounts of why induction fails
    where deduction succeeds, and about complementary duals, and the
    great account of structure and geometry and arithmetic, and continuity >>>> and infinity, up-front.

    i have a hard time parsing what u mean even with gemenigpts help, but
    please do remember the /theory of computing/ is inherently limited to
    the cardinality of natural numbers, as mechanical computations must be
    mapped to the finite machine description which compute them


    Then, that there are more models of laws of large numbers, and
    that naive induction is basically finitistic and only knows a
    law of small numbers, or "weak" induction, compared to the "strong"
    induction which is given an account by the existence of a space
    and its structure, that the "a priori" of "strong induction" is
    due classical-expositions of super-classical results after the
    double-reductio and the ad-infinitum instead of the ad-absurdam,
    about the "ab-absurdam", that numbers and forms always have those.


    Then, retro-finitism or ultra-finitism is naive, and incomplete.


    Hm. Thanks for writing.

    Agreeably, accounts of the _unbounded_, of the _finite_, keep
    things "sensible, fungible, and tractable", and that accounts
    of _completions_, in the _infinite_, are due descriptions by
    super-classical results like what give the geometric series,
    or Zeno's arguments either way _both existing_ when induction
    either way makes a counter-development/counter-example to the other,
    this is about "weak and strong induction", that induction and infinite
    induction it's usual given account of the base case then inductive case,
    is "weak", since it's _completion_, doesn't have an "actual
    infinite".


    For the theory of computation, a usual idea is that there are
    "Turing machines" and then "Zeno machines", then, that like
    there are the "digital" and "analog", the "discrete" and "continuous",
    that it's independent number theory which of large, larger, and largest
    laws of large numbers apply, for inductive, infinite, and continuum
    limits.

    There are at least three kinds of limits, then that if infinite limits
    and continuum limits require the "strong induction", i.e., weak
    induction plus also another reason why the completion occurs,
    then makes for "completeness" and "measure" after "density".

    The unbounded readily gives density, it's agreeable.


    I suppose it's been called "non-standard", like "non-standard models
    of integers" or "non-standard models of probability", yet, that's
    because the entire linear curriculum mostly doesn't have an account
    of deductive analysis at all, since, trivially enough, it's provides
    a direct and immediate counter-development and counter-example to
    any plain course of the naive (or weak) induction at all, which has
    been known since antiquity about why the ancients said truth was
    "discovered" instead of "invented".

    Approximation algorithms abound, to be sure, and approximations
    always have a nominally non-zero error term, and modeling the error term
    or modeling the error bounds of approximations, include for
    where the error term gets away from the approximation itself.


    Then, the "as-if" or "almost", like the "almost-all",
    "almost-everywhere", "almost-periodic", "almost-analytic", the sorts
    of "almost-approximate", that's naturally enough more of a
    "what-if" than an "as-if".


    So, "modeling the error bounds" is usually what's involved in
    analysis since here the study of things is called "dynamical
    modeling" not "chaos theory", since math is not indeterministic,
    then about singularity theory that singularities in a singularity
    theory are branches in a multiplicity theory, just taking the
    extra-book-keeping when there's enough scratch-space to write
    out the numerical method or approximative algorithm, or making
    an account of how mathematics does that in actual numerical resources,
    or for the usual accounts of incompleteness (or, inconsistency if
    you'd rather not have incompleteness) of the ordinary, and somehow
    the completeness and the consistency in the extra-ordinary,
    of the infinite and continuous.

    So, Church-Turing thesis is prolly not fritzed, though that
    it is though merely a "fragment" or the "ordinary", and,

    ross, i'm not using a model of hypercomputation to usurp the ct thesis.
    as we don't actually know how to mechanically implement those, they
    cannot be used by a human to compute something a turing machine cannot,
    and therefore do not refute it.

    recall that the ct thesis is thus:

    -a| A function is effectively calculable by a human being
    -a| _iff_ it can be computed by a turing machine.

    the ct thesis claims actual mechanical computation that a human can do
    is limited to what can be expressed within the turing machine model. refuting the ct thesis involves demonstrating a mechanical process that utilizes an idealized human agent to compute something which a turing machine simply cannot express.

    So, why not use my earlier example of the /Antikythera mechanism/ as a
    counter example. If I get it right, it's an actual mechanical calcula-
    tor, and as such, cannot be implemented in a /turing machine/ except as
    a simulation of such.

    Now, simulation isn't reality. /The map is not the territory/. I know mathematical types have a hard time with this concept, so bear with me
    for now if you're confused.

    I'm also confused about the theoretical argument about what is and isn't calculable by a turing machine. I live in the real world, and I use
    real machines.

    Basically, what I'm getting at, a human operating the /Antikythera
    mechanism/ is indeed calculating something that /cannot/ be done by
    a turing machine, and therefore refutes the /ct thesis/ above.

    Now, I know proof by counter example is not pleasant to experience, so
    I don't begrudge you being upset, but please find another topic for a
    thesis, or just present the /Antikypthera mechanism/ as a counter ex-
    ample, and be done with it.

    the refutation is not algorithmically novel or hard, it's more due to
    the fact that the human operation exists outside what can be directly referenced by the turing machine model. yes, we can simulate the human operation within the turing machine model, and that simulation will be limited by what turing machines can express, but a human agent
    mechanically doing that same thing just is not. i'm sure u won't be convinced by my paragraphs here, the full argument will be in a paper
    i'm finishing up.

    the refutation is again, not algorithmically hard, it's moreso tied to
    the same self-referential paradox which stumped turing into establishing limits to turing machine computation in the first place, back on the
    1936 paper /on computable numbers/

    in the extra-ordinary then thusly, it's _independent_, the ordinary.
    (The "standard" in one sense of the ordinary is "non-standard"
    the other, like "standard infinitesimals" or "standard integers",
    one giving a clock-arithmetic the other a field-arithmetic,
    and only agreeing about the measure of [0,1].)

    I didn't graduate as a /math major/ so I don't move in the infinite
    social circles.


    Have a nice day!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Ross Finlayson@ross.a.finlayson@gmail.com to comp.theory on Sun Aug 16 17:43:10 2026
    From Newsgroup: comp.theory

    On 08/16/2026 03:23 PM, Johann 'Myrkraverk' Oskarsson wrote:
    On 16/08/2026 5:38 AM, dart200 wrote:
    On 8/14/26 7:27 AM, Ross Finlayson wrote:
    On 08/13/2026 06:29 PM, dart200 wrote:
    On 8/13/26 10:12 AM, Ross Finlayson wrote:
    On 08/12/2026 10:15 PM, dart200 wrote:
    On 8/12/26 7:34 PM, Ross Finlayson wrote:
    On 08/07/2026 10:18 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 06/08/2026 5:29 AM, Dude wrote:

    dart200 <user7160@newsgrouper.org.invalid> posted:

    upon picking apart the abject nonsense that is recursive
    undecidability,
    so that i can discuss what undecidability _actually_ looks like >>>>>>>>>> within
    computing,

    i find myself stumbling into a proof that turing machines as a >>>>>>>>>> model are
    not capable of computing everything that can be computed
    mechanically


    Have you considered how to model the /Antikythera Mechanism/ on a >>>>>>>> Turing
    machine, as a subject for your thesis? I believe that's been
    done, so
    the question is, what exactly is it about mechanical computing that >>>>>>>> cannot be done on Turing machines?


    i guess we'll see where the paper ultimately leads for sure, but >>>>>>>>>> things
    are cooking rLiN+A

    You can't do much advanced computing on an Apple laptop operating >>>>>>>>> from
    your
    kitchen table.

    That is a matter of perspective. The fruit laptop on the kitchen >>>>>>>> table
    can be thought of as a terminal, or even just a typewriter, for the >>>>>>>> Cray.


    As an IT professional, I can say without the least hesitation, >>>>>>>>> there's
    not
    many machines that can compare with a Cray.


    I'm an I.T. professional who's never used a Cray, so I can't
    comment on
    that.


    Cray Computers are optimized for massive mathematical arrays >>>>>>>>> rather
    than
    standard scalar math. Crays are heavily utilized in nuclear
    research,
    aerospace design, codebreaking, and global satellite weather >>>>>>>>> prediction.


    So if I understand Finseth's book about how to code an Emacs
    correctly,
    a Cray supercomputer would be massively hampered were you to
    implement a
    text editor on one.

    So, which text editor do yo prefer, when coding a Cray?

    Maybe if you read Finsler and Boffa then it would be more clear
    why Church-Turing thesis is in a model of computation that simply >>>>>>> doesn't necessarily include "Zeno machines" and the like, while
    it's so for the finite and bounded, in the unbounded.

    Finsler and Boffa make some various accounts of "infinity"
    and the "extra-ordinary" that numbers naturally have that
    are usually ignored or plain banned in "ordinary" models.


    theoretical mathematician do have a tendency to try wacky shit
    when it
    comes infinite ...

    turing for example tried to extend logical systems into transfinite >>>>>> ordinals in an attempt to get around incompleteness. no idea what
    kind
    of axiom would realistically come past an infinite amount of them, >>>>>> but
    this unfortunately didn't work so it doesn't really matter eh?

    i plan to propose a limit to the incompleteness within computing,
    which
    in turn may in fact make it complete. or at least trivialize
    incompleteness to the point of irrelevancy


    Why would you do that?

    because erroneously proving a limit that does not actually exist will
    artificially limit our potential to wield computing as an applied
    technique

    like specifically when it comes to our practical application of
    computing, we do not prove what our computations do. we might test
    various inputs/outputs combos, but this is quite a bit inferior to
    proving semantics across their entire input possibility space (and not >>>> just brute forcing that proof)


    It's kind of like driving a car, and knowing the limits,
    since one never knows the limits, yet is always finding the limits,
    incompleteness is simply beyond a limit, and not being cognizant
    of it, if not quite sure where it is, risks hitting its wall,
    or driving over its cliff.

    i'm not sure how framing currently intractable problems as actually
    tractable (without limiting the domain!) risks driving over some
    kind of
    existential cliff in this case

    undecidability within computing is entirely founded in the problems of >>>> computing the semantic properties of computations, unlocking those
    kinds
    of problems as actually tractable seems like possibly a huge boon to
    the
    advancement of computing


    It's like the very notion of analysis, making
    restriction-of-comprehension keeps some things simple, other things
    un-available to reason. Somebody else needn't adopt that restriction, >>>>> and then it's like a rooster locking itself in the coop.

    i'm not proposing a reduction in power to computing, if anything this
    would be an expansion in descriptive power


    Mirimanoff points out that a model of finite ordinals is extra-
    ordinary,
    Russell wishes it away, the "Russell-ian retro-thesis",
    it's natural that infinity is "in", and that models of ordinals
    or integers start as both bounded-fragments and
    extraordinary-extensions, the "standard" model not even existing
    except as
    a limit of those.

    The classical expositions of the super-classical, usually considered >>>>> to start with Zeno's account of infinite-divisbility and summability >>>>> and motion in time, and the geometric series, gives reasons why that >>>>> for infinitary-analysis, that's about the only place to start that
    isn't wacky, and it makes of itself accounts of why induction fails
    where deduction succeeds, and about complementary duals, and the
    great account of structure and geometry and arithmetic, and continuity >>>>> and infinity, up-front.

    i have a hard time parsing what u mean even with gemenigpts help, but
    please do remember the /theory of computing/ is inherently limited to
    the cardinality of natural numbers, as mechanical computations must be >>>> mapped to the finite machine description which compute them


    Then, that there are more models of laws of large numbers, and
    that naive induction is basically finitistic and only knows a
    law of small numbers, or "weak" induction, compared to the "strong"
    induction which is given an account by the existence of a space
    and its structure, that the "a priori" of "strong induction" is
    due classical-expositions of super-classical results after the
    double-reductio and the ad-infinitum instead of the ad-absurdam,
    about the "ab-absurdam", that numbers and forms always have those.


    Then, retro-finitism or ultra-finitism is naive, and incomplete.


    Hm. Thanks for writing.

    Agreeably, accounts of the _unbounded_, of the _finite_, keep
    things "sensible, fungible, and tractable", and that accounts
    of _completions_, in the _infinite_, are due descriptions by
    super-classical results like what give the geometric series,
    or Zeno's arguments either way _both existing_ when induction
    either way makes a counter-development/counter-example to the other,
    this is about "weak and strong induction", that induction and infinite
    induction it's usual given account of the base case then inductive case, >>> is "weak", since it's _completion_, doesn't have an "actual
    infinite".


    For the theory of computation, a usual idea is that there are
    "Turing machines" and then "Zeno machines", then, that like
    there are the "digital" and "analog", the "discrete" and "continuous",
    that it's independent number theory which of large, larger, and largest
    laws of large numbers apply, for inductive, infinite, and continuum
    limits.

    There are at least three kinds of limits, then that if infinite limits
    and continuum limits require the "strong induction", i.e., weak
    induction plus also another reason why the completion occurs,
    then makes for "completeness" and "measure" after "density".

    The unbounded readily gives density, it's agreeable.


    I suppose it's been called "non-standard", like "non-standard models
    of integers" or "non-standard models of probability", yet, that's
    because the entire linear curriculum mostly doesn't have an account
    of deductive analysis at all, since, trivially enough, it's provides
    a direct and immediate counter-development and counter-example to
    any plain course of the naive (or weak) induction at all, which has
    been known since antiquity about why the ancients said truth was
    "discovered" instead of "invented".

    Approximation algorithms abound, to be sure, and approximations
    always have a nominally non-zero error term, and modeling the error term >>> or modeling the error bounds of approximations, include for
    where the error term gets away from the approximation itself.


    Then, the "as-if" or "almost", like the "almost-all",
    "almost-everywhere", "almost-periodic", "almost-analytic", the sorts
    of "almost-approximate", that's naturally enough more of a
    "what-if" than an "as-if".


    So, "modeling the error bounds" is usually what's involved in
    analysis since here the study of things is called "dynamical
    modeling" not "chaos theory", since math is not indeterministic,
    then about singularity theory that singularities in a singularity
    theory are branches in a multiplicity theory, just taking the
    extra-book-keeping when there's enough scratch-space to write
    out the numerical method or approximative algorithm, or making
    an account of how mathematics does that in actual numerical resources,
    or for the usual accounts of incompleteness (or, inconsistency if
    you'd rather not have incompleteness) of the ordinary, and somehow
    the completeness and the consistency in the extra-ordinary,
    of the infinite and continuous.

    So, Church-Turing thesis is prolly not fritzed, though that
    it is though merely a "fragment" or the "ordinary", and,

    ross, i'm not using a model of hypercomputation to usurp the ct
    thesis. as we don't actually know how to mechanically implement those,
    they cannot be used by a human to compute something a turing machine
    cannot, and therefore do not refute it.

    recall that the ct thesis is thus:

    | A function is effectively calculable by a human being
    | _iff_ it can be computed by a turing machine.

    the ct thesis claims actual mechanical computation that a human can do
    is limited to what can be expressed within the turing machine model.
    refuting the ct thesis involves demonstrating a mechanical process
    that utilizes an idealized human agent to compute something which a
    turing machine simply cannot express.

    So, why not use my earlier example of the /Antikythera mechanism/ as a counter example. If I get it right, it's an actual mechanical calcula-
    tor, and as such, cannot be implemented in a /turing machine/ except as
    a simulation of such.

    Now, simulation isn't reality. /The map is not the territory/. I know mathematical types have a hard time with this concept, so bear with me
    for now if you're confused.

    I'm also confused about the theoretical argument about what is and isn't calculable by a turing machine. I live in the real world, and I use
    real machines.

    Basically, what I'm getting at, a human operating the /Antikythera
    mechanism/ is indeed calculating something that /cannot/ be done by
    a turing machine, and therefore refutes the /ct thesis/ above.

    Now, I know proof by counter example is not pleasant to experience, so
    I don't begrudge you being upset, but please find another topic for a
    thesis, or just present the /Antikypthera mechanism/ as a counter ex-
    ample, and be done with it.

    the refutation is not algorithmically novel or hard, it's more due to
    the fact that the human operation exists outside what can be directly
    referenced by the turing machine model. yes, we can simulate the human
    operation within the turing machine model, and that simulation will be
    limited by what turing machines can express, but a human agent
    mechanically doing that same thing just is not. i'm sure u won't be
    convinced by my paragraphs here, the full argument will be in a paper
    i'm finishing up.

    the refutation is again, not algorithmically hard, it's moreso tied to
    the same self-referential paradox which stumped turing into
    establishing limits to turing machine computation in the first place,
    back on the 1936 paper /on computable numbers/

    in the extra-ordinary then thusly, it's _independent_, the ordinary.
    (The "standard" in one sense of the ordinary is "non-standard"
    the other, like "standard infinitesimals" or "standard integers",
    one giving a clock-arithmetic the other a field-arithmetic,
    and only agreeing about the measure of [0,1].)

    I didn't graduate as a /math major/ so I don't move in the infinite
    social circles.


    Have a nice day!




    Somebody like Maugin has something like the "acoustic wave transducer",
    an electronic circuit that makes for solving what would be digital
    problems more like analog problems.

    These days they call that a "D-Wave quantum simulator computer",
    about that there are then also many models of computation that
    were about planar circuits, with vias and all, in the "three-dimensional integrated circuit" or "3-D IC", that make some
    what were fundamentally serial algorithms into general parallel.

    The use of optical junctions and optical traces instead of electronic
    junctions and electronic traces, offers an entirely different model
    of propagation, basically for 1-to-many or many-to-1.


    It's agreeable that a usual model of computation as a calculus
    is limited to an abacus (if a rather large abacus): yet accounts
    like the integral or infinitesimal analysis or Fourier-style analysis
    or about doubling-measures and quasi-invariant measure theory, these
    are "super-classical results" then though that a "computer algebra
    system" can be given the rules to work on them like perfect numbers
    in their symbolic forms.

    Then, figuring whether finite, if large, approximations are "close
    enough" that the differences would be lost, the "negligeable",
    also they're formally "not quite", about what's called "weak induction",
    and that the definition of "strong induction" is
    that some super-classical example that's somehow geometric and
    infinitary already exists, ..., there may be many practical and
    successful "polynomial-time approximations to NP-hard problems",
    and in fact there's a great book with that title, yet otherwise
    the "Turing-complete" as the same as "induction-complete" isn't so.


    Mathematics is kind of like physics, the popular accounts:
    rife with incompleteness, and inconsistencies, "independence",
    that various successful and practical accounts choose to ignore
    (and others just don't even know, nor, as long as they're "close
    enough", care).






    --- Synchronet 3.22a-Linux NewsLink 1.2