Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail From: Mikko Newsgroups: comp.theory Subject: Re: Every D(D) simulated by H presents non-halting behavior to H ### Date: Mon, 27 May 2024 11:11:40 +0300 Organization: - Lines: 169 Message-ID: References: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 8bit Injection-Date: Mon, 27 May 2024 10:11:40 +0200 (CEST) Injection-Info: dont-email.me; posting-host="15e20b36553d1d7cbc0e38273b457200"; logging-data="4149733"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1/ux2Efy3DIlGY/4/MynT0C" User-Agent: Unison/2.2 Cancel-Lock: sha1:OvWJrpIE+FMF2RrAcW1ca2eX3hI= Bytes: 9275 On 2024-05-26 16:50:21 +0000, olcott said: > On 5/26/2024 11:19 AM, Mikko wrote: >> On 2024-05-26 13:32:08 +0000, olcott said: >> >>> On 5/26/2024 3:57 AM, Mikko wrote: >>>> On 2024-05-25 12:09:18 +0000, olcott said: >>>> >>>>> On 5/25/2024 3:14 AM, Mikko wrote: >>>>>> On 2024-05-24 17:13:05 +0000, olcott said: >>>>>> >>>>>>> On 5/24/2024 3:58 AM, Mikko wrote: >>>>>>>> On 2024-05-23 13:18:02 +0000, olcott said: >>>>>>>> >>>>>>>>> On 5/23/2024 5:06 AM, Mikko wrote: >>>>>>>>>> On 2024-05-22 14:51:50 +0000, olcott said: >>>>>>>>>> >>>>>>>>>>> On 5/22/2024 2:39 AM, Mikko wrote: >>>>>>>>>>>> On 2024-05-21 13:54:09 +0000, olcott said: >>>>>>>>>>>>> You are asking for the definition of correct simulation >>>>>>>>>>>>> that I have been providing for quite a while recently. >>>>>>>>>>>> >>>>>>>>>>>> That was not my main intent. I wanted to know why your >>>>>>>>>>>> statement >>>>>>>>>>>> >>>>>>>>>>>>>>>>>>> No D simulated correctly by any H of every H/D pair specified >>>>>>>>>>>>>>>>>>> by the above template ever reaches its own line 06 and halts. >>>>>>>>>>>> >>>>>>>>>>>> exludes every unsimulated or incorrectly simulated D? >>>>>>>>>>> >>>>>>>>>>> That sounds like Richard that assumed that incorrect answers are OK >>>>>>>>>>> unless I specifically say that incorrect answers are not OK. >>>>>>>>>> >>>>>>>>>> Maybe but I don't promise that the response to the incorrect answer >>>>>>>>>> will sound the same. >>>>>>>>>> >>>>>>>>>>> On 5/19/2024 12:17 PM, Richard Damon wrote: >>>>>>>>>>>  > On 5/19/24 9:59 AM, olcott wrote: >>>>>>>>>>>  >> Richard has stated that he thinks that an example of >>>>>>>>>>>  >> {D never simulated by H} ∈ {every D simulated by H} >>>>>>>>>>>  > >>>>>>>>>>>  > No, the H that didn't simulate its input shows that >>>>>>>>>>>  > *once you allow H to not be required to be correct*, >>>>>>>>>>>  > that we can then have a trivial function that is >>>>>>>>>>>  > "just as correct" (since wrong answers were allowed). >>>>>>>>>>> >>>>>>>>>>>>> A c function is correctly simulated when its machine language >>>>>>>>>>>>> instructions are emulated with an x86 emulator in the order >>>>>>>>>>>>> that they are specified by the x86 machine language of this >>>>>>>>>>>>> c function. >>>>>>>>>>>> >>>>>>>>>>>> Does "its machine language instructions" mean all executed instructions >>>>>>>>>>>> until the progam terminates? Or from the start of the program until >>>>>>>>>>>> there is no reason to continue? Or from some point to some other point? >>>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> It means that 1 to N instructions of D are correctly simulated >>>>>>>>>>> by pure function H. Because D correctly simulated by H remains >>>>>>>>>>> stuck in recursive simulation D cannot possibly reach is own >>>>>>>>>>> line 06 and halt. >>>>>>>>>> >>>>>>>>>> If you mean that H cannot simulate D to the line 06 then say so. >>>>>>>>>> A D that is simulated by H is D and so is a D that is not simulated >>>>>>>>>> by H so both can do what a D can do. Saying "simulated by H" adds >>>>>>>>>> nothing. >>>>>>>>>> >>>>>>>>>>>>> For non-terminating functions we can only correctly >>>>>>>>>>>>> simulate N machine language instructions. >>>>>>>>>>>> >>>>>>>>>>>> But does you definition regard that partial simulation as "correct >>>>>>>>>>>> simulation"? >>>>>>>>>>> >>>>>>>>>>> When 1 to 2^64 instructions of D are correctly simulated by H >>>>>>>>>>> it becomes clear that for every H/D pair of the infinite set >>>>>>>>>>> of H/D pairs D correctly simulated by H remains stuck in recursive >>>>>>>>>>> simulation. >>>>>>>>>> >>>>>>>>>> If you think that the meaning of "correctly simulate" is not >>>>>>>>>> important you should not use those words. >>>>>>>>>> >>>>>>>>> >>>>>>>>> I must use those words or a standard of incorrect simulation >>>>>>>>> is assumed. >>>>>>>> >>>>>>>> There is no "standard of incorrect simulation". >>>>>>>> >>>>>>>>> We have been going over the term "correct simulation" >>>>>>>>> in these forums with dozens of people and hundreds of messages >>>>>>>>> over several years. >>>>>>>> >>>>>>>> That alone is a sufficient reaston to avoid the expression. >>>>>>>> >>>>>>>>> CORRECT SIMULATION DEFINED >>>>>>>>>    In the above case a simulator is an x86 emulator that correctly >>>>>>>>>    emulates at least one of the x86 instructions of D in the order >>>>>>>>>    specified by the x86 instructions of D. >>>>>>>>> >>>>>>>>>    This may include correctly emulating the x86 instructions of H in the >>>>>>>>>    order specified by the x86 instructions of H thus calling H(D,D) in >>>>>>>>>    recursive simulation. >>>>>>>> >>>>>>>> That is not a definition but perhaps a suffient substitute for paractical >>>>>>>> purposes. >>>>>>>> >>>>>>> >>>>>>> It provides a clear and correct criterion measure to utterly >>>>>>> refute each and every reviewer that tries to get away with >>>>>>> the incorrect emulation of the x86 instructions of H or D or >>>>>>> emulating them in the wrong order. >>>>>> >>>>>> You may call it a "diagnostic criterion" or just a "criterion" but >>>>>> it does not define anything. Whether it is clear or sufficient is >>>>>> another problem. >>>>>> >>>>> >>>>> For over two years I had two dozen people unified in consensus >>>>> continue to insist that a correct simulation of D by H did not >>>>> require emulating the x86 machine language instructions of D >>>>> correctly or in the correct order specified by D. >>>> >>>> Is the disagreement about the meaning of "correct" or "simulation" >>>> or some other word, or is the disagreement about correctness of the >>>> simulation? >>>> >>> >>>     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. >>> >>> People disagree with the above definition. >> >> Then you may ask what other word would be better instead of "simulator". >> >>> They believe that a correct >>> simulation requires H to report on the computation that itself is >>> contained within: >> >> Why should the term "simulation" imply anything about reporting? >> > > There you go an actual mistake that I made. > I will phrase what I mean more accurately. > > Everyone reviewing my work agrees that D correctly simulated by H should > simulate the behavior of the directly executed D(D) thus not the actual > behavior of D correctly simulated by pure function H. The part after "thus" it not necessary. It merely comments whether H should do what it does, which it does anyway. > When we see that D correctly simulated by pure simulator H would remain > stuck in recursive simulation then we also know that D never reaches its > own line 06 and halts in less than an infinite number of correctly > simulated steps. Which means that H never terminates. You said that by your definition a function that never terminates is not a pure function. Therefore H, if it exists, is not a pure function, and the phrase "pure function H" does not denote. > This means that D correctly simulated by pure function H also never > reaches it own line 06 and halts. Yes, if H never terminates then neither does D. -- Mikko