Path: ...!2.eu.feeder.erje.net!feeder.erje.net!news.mb-net.net!open-news-network.org!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail From: olcott Newsgroups: comp.theory,sci.logic Subject: =?UTF-8?Q?Re=3A_A_simulating_halt_decider_applied_to_the_The_Peter_?= =?UTF-8?Q?Linz_Turing_Machine_description_=E2=9F=A8=C4=A4=E2=9F=A9?= Date: Mon, 27 May 2024 19:26:45 -0500 Organization: A noiseless patient Spider Lines: 148 Message-ID: References: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Injection-Date: Tue, 28 May 2024 02:26:46 +0200 (CEST) Injection-Info: dont-email.me; posting-host="62ab2bf33c274f123184493b42753dfc"; logging-data="299528"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX18qunyZp9CcdUxmbJ7yhxOp" User-Agent: Mozilla Thunderbird Cancel-Lock: sha1:qdZJdW1GQTCRd/U6zSMV/98ecl8= Content-Language: en-US In-Reply-To: Bytes: 8142 On 5/27/2024 7:17 PM, Richard Damon wrote: > On 5/27/24 8:08 PM, olcott wrote: >> On 5/27/2024 5:44 PM, Richard Damon wrote: >>> On 5/27/24 6:32 PM, olcott wrote: >>>> On 5/27/2024 4:21 PM, Richard Damon wrote: >>>>> On 5/27/24 3:45 PM, olcott wrote: >>>>>> On 5/27/2024 11:33 AM, Richard Damon wrote: >>>>>>> On 5/27/24 12:22 PM, olcott wrote: >>>>>>>> On 5/27/2024 10:58 AM, Richard Damon wrote: >>>>>>>>> On 5/27/24 11:46 AM, olcott wrote: >>>>>>>>>> On 5/27/2024 10:25 AM, Richard Damon wrote: >>>>>>>>>>> On 5/27/24 11:06 AM, olcott wrote: >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> typedef int (*ptr)();  // ptr is pointer to int function in C >>>>>>>>>> 00       int H(ptr p, ptr i); >>>>>>>>>> 01       int D(ptr p) >>>>>>>>>> 02       { >>>>>>>>>> 03         int Halt_Status = H(p, p); >>>>>>>>>> 04         if (Halt_Status) >>>>>>>>>> 05           HERE: goto HERE; >>>>>>>>>> 06         return Halt_Status; >>>>>>>>>> 07       } >>>>>>>>>> 08 >>>>>>>>>> 09       int main() >>>>>>>>>> 10       { >>>>>>>>>> 11         H(D,D); >>>>>>>>>> 12         return 0; >>>>>>>>>> 13       } >>>>>>>>>> >>>>>>>>>> The above template refers to an infinite set of H/D pairs >>>>>>>>>> where D is >>>>>>>>>> correctly simulated by either pure simulator H or pure >>>>>>>>>> function H. This >>>>>>>>>> was done because many reviewers used the shell game ploy to >>>>>>>>>> endlessly >>>>>>>>>> switch which H/D pair was being referred to. >>>>>>>>>> >>>>>>>>>> *Correct Simulation Defined* >>>>>>>>>>     This is provided because many reviewers had a different >>>>>>>>>> notion of >>>>>>>>>>     correct simulation that diverges from this notion. >>>>>>>>>> >>>>>>>>>>     A simulator is an x86 emulator that correctly emulates 1 >>>>>>>>>> to N of the >>>>>>>>>>     x86 instructions of D in the order specified by the x86 >>>>>>>>>> instructions >>>>>>>>>>     of D. This may include M recursive emulations of H >>>>>>>>>> emulating itself >>>>>>>>>>     emulating D. >>>>>>>>> >>>>>>>>> And how do you apply that to a TEMPLATE that doesn't define >>>>>>>>> what a call H means (as it could be any of the infinite set of >>>>>>>>> Hs that you can instantiate the template on)? >>>>>>>>> >>>>>>>> >>>>>>>> *Somehow we got off track of the subject of this thread* >>>>>>> >>>>>>> I note that YOU keep on switching between your C program and >>>>>>> Turing Machines. >>>>>>> >>>>>>> Note, per the implications that you implicitly agreed to (by not >>>>>>> even trying to refute) the two systems are NOT equivalents of >>>>>>> each other. >>>>>>> >>>>>> >>>>>> (1) I think you are wrong. I have not seen any of your >>>>>> reasoning that was not anchored in false assumptions. >>>>>> Your make fake rebuttal is to change the subject. >>>>>> >>>>>> (2) It does not matter my proof is anchored in the Linz >>>>>> proof and the H/D pairs are only used to have a 100% concrete >>>>>> basis to perfectly anchor things such as the correct meaning >>>>>> of D correctly simulated by H so that people cannot get away >>>>>> with claiming that an incorrect simulation is correct. >>>>>> >>>>>> int main() { D(D); } IS NOT THE BEHAVIOR OF D CORRECTLY SIMULATED >>>>>> BY H. >>>>>> One cannot simply ignore the pathological relationship between H >>>>>> and D. >>>>>> >>>>>>>> >>>>>>>> When Ĥ is applied to ⟨Ĥ⟩ >>>>>>>> Ĥ.q0 ⟨Ĥ⟩ ⊢* embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ ⊢* Ĥ.qy ∞ >>>>>>>> Ĥ.q0 ⟨Ĥ⟩ ⊢* embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ ⊢* Ĥ.qn >>>>>>>> >>>>>>>>   Ĥ copies its own Turing machine description: ⟨Ĥ⟩ >>>>>>>>   then invokes embedded_H that simulates ⟨Ĥ⟩ with ⟨Ĥ⟩ as input. >>>>>>>> >>>>>>>> For the purposes of the above analysis we hypothesize that >>>>>>>> embedded_H is either a UTM or a UTM that has been adapted >>>>>>>> to stop simulating after a finite number of steps of simulation. >>>>>>> >>>>>>> And what you do mean by that? >>>>>>> >>>>>>> Do you hypothesize that the original H was just a pure UTM, >>>>>> >>>>>> The original proof does not consider the notion of a simulating >>>>>> halt decider so I have to begin the proof at an earlier stage >>>>>> than any definition of H. >>>>> >>>>> The biggest problem is that the input to the Turing machine decider >>>>> H is the description of a Turing Machine H^, which is a SPECIFIC >>>>> machine, >>>> >>>> When you say "specific machine" you don't mean anything like a >>>> 100% completely specified sequence of state transitions encoded >>>> as a single unique finite string. >>> >>> Mostly. >>> >>> There doesn't need to be a unique finite string, but it is a 100% >>> completely specified state transition/tape operation table. >>> >> >> When Ĥ is applied to ⟨Ĥ⟩ >> Ĥ.q0 ⟨Ĥ⟩ ⊢* embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ ⊢* Ĥ.qy ∞ >> Ĥ.q0 ⟨Ĥ⟩ ⊢* embedded_H ⟨Ĥ⟩ ⟨Ĥ⟩ ⊢* Ĥ.qn >> >> In other words Linz did not prove that there are no set >> of state transitions specified by ⊢* that derives the >> correct halt status of ⟨Ĥ⟩ ⟨Ĥ⟩. >> >> He only said there there is one specific machine that >> gets the wrong answer. >> > > He STARTS with a proof that one specific (but arbitrary) machine gets > the wrong answer. > > Then he shows that the same proof can be applied to ANY such machine > (becaue the proof didn't depend on any specific details of the machine, > just the general properties of that machine) > > I guess you don't understand how to do categorical proofs. > I totally do. Can you please write down the "completely specified state transition/tape operation table." of this specific (thus uniquely identifiable) machine I would really like to see it. -- Copyright 2024 Olcott "Talent hits a target no one else can hit; Genius hits a target no one else can see." Arthur Schopenhauer