Path: news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail From: dbush Newsgroups: comp.lang.c Subject: Re: DD simulated by HHH cannot possibly halt (Halting Problem) Date: Sat, 5 Apr 2025 12:25:12 -0400 Organization: A noiseless patient Spider Lines: 128 Message-ID: References: MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Injection-Date: Sat, 05 Apr 2025 18:25:12 +0200 (CEST) Injection-Info: dont-email.me; posting-host="43eaa133a904c8986c7a0672d25633a8"; logging-data="2851472"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1+w58IR8dNBRefnNgmzNF/K" User-Agent: Mozilla Thunderbird Cancel-Lock: sha1:l+6RJT6xaAWb2b67f6TDXMQK9xY= Content-Language: en-US In-Reply-To: On 4/5/2025 11:59 AM, olcott wrote: > On 4/5/2025 2:42 AM, Richard Heathfield wrote: >> On 05/04/2025 07:14, olcott wrote: >>> On 4/4/2025 10:49 PM, Richard Heathfield wrote: >>>> On 05/04/2025 00:41, olcott wrote: >>>>> *Simulating termination analyzer Principle* >>>>> It is always correct for any simulating termination >>>>> analyzer to stop simulating and reject any input that >>>>> would otherwise prevent its own termination. The >>>>> only rebuttal to this is rejecting the notion that >>>>> deciders must always halt. >>>> >>> >>> typedef void (*ptr)(); >>> int HHH(ptr P); >>> >>> int DD() >>> { >>>    int Halt_Status = HHH(DD); >>>    if (Halt_Status) >>>      HERE: goto HERE; >>>    return Halt_Status; >>> } >>> >>> int main() >>> { >>>    HHH(DD); >>> } >>> >>>> In other words, you operate on the principle that deciders don't >>>> have to (and indeed can't) always make a correct decision on whether >>>> an input program halts. >>>> >>> >>> The termination analyzer HHH would be correct >>> to determine that it must stop simulating DD to >>> prevent its own non-termination >> >> Fine, but then it fails to do its job. What you are learning (albeit >> slowly) is that the termination analyser HHH can't analyse whether DD >> terminates. It is therefore not a general purpose termination analyser. >> > > Introduction to the Theory of Computation 3rd Edition > by Michael Sipser (Author) (best selling textbook) > > >     If simulating halt decider H correctly simulates its input D >     until H correctly determines that its simulated D would never >     stop running unless aborted then > >     H can abort its simulation of D and correctly report that D >     specifies a non-halting sequence of configurations. > But not what you think he agreed to: On Monday, March 6, 2023 at 2:41:27 PM UTC-5, Ben Bacarisse wrote: > Fritz Feldhase writes: > > > On Monday, March 6, 2023 at 3:56:52 AM UTC+1, olcott wrote: > >> On 3/5/2023 8:33 PM, Fritz Feldhase wrote: > >> > On Monday, March 6, 2023 at 3:30:38 AM UTC+1, olcott wrote: > >> > > > >> > > I needed Sipser for people [bla] > >> > > > >> > Does Sipser support your view/claim that you have refuted the halting theorem? > >> > > >> > Does he write/teach that the halting theorem is invalid? > >> > > >> > Tell us, oh genius! > >> > > >> Professor Sipser only agreed that [...] > > > > So the answer is no. Noted. > > > >> Because he has >250 students he did not have time to examine anything > >> else. [...] > > > > Oh, a CS professor does not have the time to check a refutation of the > > halting theorem. *lol* > I exchanged emails with him about this. He does not agree with anything > substantive that PO has written. I won't quote him, as I don't have > permission, but he was, let's say... forthright, in his reply to me. > On 8/23/2024 5:07 PM, Ben Bacarisse wrote: > joes writes: > >> Am Wed, 21 Aug 2024 20:55:52 -0500 schrieb olcott: > >>> Professor Sipser clearly agreed that an H that does a finite simulation >>> of D is to predict the behavior of an unlimited simulation of D. >> >> If the simulator *itself* would not abort. The H called by D is, >> by construction, the same and *does* abort. > > We don't really know what context Sipser was given. I got in touch at > the time so do I know he had enough context to know that PO's ideas were > "wacky" and that had agreed to what he considered a "minor remark". > > Since PO considers his words finely crafted and key to his so-called > work I think it's clear that Sipser did not take the "minor remark" he > agreed to to mean what PO takes it to mean! My own take if that he > (Sipser) read it as a general remark about how to determine some cases, > i.e. that D names an input that H can partially simulate to determine > it's halting or otherwise. We all know or could construct some such > cases. > > I suspect he was tricked because PO used H and D as the names without > making it clear that D was constructed from H in the usual way (Sipser > uses H and D in at least one of his proofs). Of course, he is clued in > enough know that, if D is indeed constructed from H like that, the > "minor remark" becomes true by being a hypothetical: if the moon is made > of cheese, the Martians can look forward to a fine fondue. But, > personally, I think the professor is more straight talking than that, > and he simply took as a method that can work for some inputs. That's > the only way is could be seen as a "minor remark" with being accused of > being disingenuous. On 8/23/2024 9:10 PM, Mike Terry wrote: > So that PO will have no cause to quote me as supporting his case: what > Sipser understood he was agreeing to was NOT what PO interprets it as > meaning. Sipser would not agree that the conclusion applies in PO's > HHH(DDD) scenario, where DDD halts.