XPost: comp.theory, sci.logic, sci.math   
   XPost: comp.software-eng   
   From: polcott333@gmail.com   
      
   On 1/11/2026 4:13 AM, Mikko wrote:   
   > On 10/01/2026 17:47, olcott wrote:   
   >> On 1/10/2026 2:23 AM, Mikko wrote:   
   >>> On 09/01/2026 17:52, olcott wrote:   
   >>>> On 1/9/2026 3:59 AM, Mikko wrote:   
   >>>>> On 08/01/2026 16:22, olcott wrote:   
   >>>>>> On 1/8/2026 4:22 AM, Mikko wrote:   
   >>>>>>> On 07/01/2026 13:54, olcott wrote:   
   >>>>>>>> On 1/7/2026 5:49 AM, Mikko wrote:   
   >>>>>>>>> On 07/01/2026 06:44, olcott wrote:   
   >>>>>>>>>> All deciders essentially: Transform finite string   
   >>>>>>>>>> inputs by finite string transformation rules into   
   >>>>>>>>>> {Accept, Reject} values.   
   >>>>>>>>>>   
   >>>>>>>>>> The counter-example input to requires more than   
   >>>>>>>>>> can be derived from finite string transformation   
   >>>>>>>>>> rules applied to this specific input thus the   
   >>>>>>>>>> Halting Problem requires too much.   
   >>>>>>>>   
   >>>>>>>>> In a sense the halting problem asks too much: the problem is   
   >>>>>>>>> proven to   
   >>>>>>>>> be unsolvable. In another sense it asks too little: usually we   
   >>>>>>>>> want to   
   >>>>>>>>> know whether a method halts on every input, not just one.   
   >>>>>>>>>   
   >>>>>>>>> Although the halting problem is unsolvable, there are partial   
   >>>>>>>>> solutions   
   >>>>>>>>> to the halting problem. In particular, every counter-example to   
   >>>>>>>>> the   
   >>>>>>>>> full solution is correctly solved by some partial deciders.   
   >>>>>>>>   
   >>>>>>>> *if undecidability is correct then truth itself is broken*   
   >>>>>>>   
   >>>>>>> Depends on whether the word "truth" is interpeted in the standard   
   >>>>>>> sense or in Olcott's sense.   
   >>>>>>   
   >>>>>> Undecidability is misconception. Self-contradictory   
   >>>>>> expressions are correctly rejected as semantically   
   >>>>>> incoherent thus form no undecidability or incompleteness.   
   >>>>>   
   >>>>> The misconception is yours. No expression in the language of the first   
   >>>>> order group theory is self-contradictory. But the first order goupr   
   >>>>> theory is incomplete: it is impossible to prove that AB = BA is true   
   >>>>> for every A and every B but it is also impossible to prove that AB   
   >>>>> = BA   
   >>>>> is false for some A and some B.   
   >>>>>   
   >>>>   
   >>>> All deciders essentially: Transform finite string   
   >>>> inputs by finite string transformation rules into   
   >>>> {Accept, Reject} values.   
   >>>>   
   >>>> When a required result cannot be derived by applying   
   >>>> finite string transformation rules to actual finite   
   >>>> string inputs, then the required result exceeds the   
   >>>> scope of computation and must be rejected as an   
   >>>> incorrect requirement.   
   >>>   
   >>> No, that does not follow. If a required result cannot be derived by   
   >>> appying a finite string transformation then the it it is uncomputable.   
   >>   
   >> Right. Outside the scope of computation. Requiring anything   
   >> outside the scope of computation is an incorrect requirement.   
   >   
   > You can't determine whether the required result is computable before   
   > you have the requirement.   
   >   
      
      
   *Computation and Undecidability*   
   https://philpapers.org/go.pl?aid=OLCCAU   
      
   We know that there does not exist any finite   
   string transformations that H can apply to its   
   input P to derive the halt status of any P   
   that does the opposite of whatever H returns.   
      
   *ChatGPT explains how and why I am correct*   
      
      *Reinterpretation of undecidability*   
      The example of P and H demonstrates that what is   
      often called “undecidable” is better understood as   
      ill-posed with respect to computable semantics.   
      When the specification is constrained to properties   
      detectable via finite simulation and finite pattern   
      recognition, computation proceeds normally and   
      correctly. Undecidability only appears when the   
      specification overreaches that boundary.   
      
   Every other LLM says this same thing using   
   different words.   
      
   >>   
   >>> Of course, it one can prove that the required result is not computable   
   >>> then that helps to avoid wasting effort to try the impossible. The   
   >>> situation is worse if it is not known that the required result is not   
   >>> computable.   
   >>>   
   >>> That something is not computable does not mean that there is anyting   
   >>> "incorrect" in the requirement.   
   >>   
   >> Yes it certainly does. Requiring the impossible is always an error.   
   >> Requiring an answer to a yes/no question that has no correct yes/no   
   >> answer is an incorrect question that must be rejected.   
   >>   
   >>> In order to claim that a requirement   
   >>> is incorrect one must at least prove that the requirement does not   
   >>> serve its intended purpose.   
   >>   
   >> Requiring the impossible cannot possibly serve any purpose   
   >> except perhaps to exemplify one's own ignorance.   
   >>   
   >>> Even then it is possible that the   
   >>> requirement serves some other purpose. Even if a requirement serves   
   >>> no purpose that need not mean that it be "incorrect", only that it   
   >>> is useless.   
   >>>   
   >>   
   >>   
   >>   
   >   
   >   
      
      
   --   
   Copyright 2026 Olcott

   
      
   My 28 year goal has been to make 
   
   "true on the basis of meaning expressed in language"
   
   reliably computable.

   
      
   This required establishing a new foundation
   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)