Path: ...!weretis.net!feeder9.news.weretis.net!i2pn.org!i2pn2.org!.POSTED!not-for-mail
From: Richard Damon <richard@damon-family.org>
Newsgroups: comp.theory
Subject: Re: Can D simulated by H terminate normally?
Date: Thu, 2 May 2024 21:48:48 -0400
Organization: i2pn2 (i2pn.org)
Message-ID: <v11fq0$2tlr1$3@i2pn2.org>
References: <v0k4jc$laej$1@dont-email.me> <v0m5sn$172p4$1@dont-email.me>
 <v0oban$1o3b$1@news.muc.de> <v0oce3$1q3aq$4@dont-email.me>
 <v0oe1b$1o3b$2@news.muc.de> <v0ofl3$1r1mf$1@dont-email.me>
 <v0oh7g$1o3b$3@news.muc.de> <v0olhv$1sgeo$1@dont-email.me>
 <v0oobd$1o3b$4@news.muc.de> <v0or07$1tmga$1@dont-email.me>
 <v0qb59$2bsfc$1@dont-email.me> <v0r242$2hb7o$1@dont-email.me>
 <v0r3kh$hka$1@news.muc.de> <v0r5f2$2hb7o$11@dont-email.me>
 <v0rsbv$2m1nf$8@i2pn2.org> <v0sgcm$2varu$3@dont-email.me>
 <v0vmvu$209h$3@news.muc.de> <v10md7$ese$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Fri, 3 May 2024 01:48:48 -0000 (UTC)
Injection-Info: i2pn2.org;
	logging-data="3069793"; mail-complaints-to="usenet@i2pn2.org";
	posting-account="diqKR1lalukngNWEqoq9/uFtbkm5U+w3w6FQ0yesrXg";
User-Agent: Mozilla Thunderbird
X-Spam-Checker-Version: SpamAssassin 4.0.0
Content-Language: en-US
In-Reply-To: <v10md7$ese$1@dont-email.me>
Bytes: 5844
Lines: 122

On 5/2/24 2:35 PM, olcott wrote:
> On 5/2/2024 4:39 AM, Alan Mackenzie wrote:
>> olcott <polcott333@gmail.com> wrote:
>>> On 4/30/2024 5:46 PM, Richard Damon wrote:
>>>> On 4/30/24 12:15 PM, olcott wrote:
>>>>> On 4/30/2024 10:44 AM, Alan Mackenzie wrote:
>>>>>> olcott <polcott333@gmail.com> wrote:
>>>>>>> On 4/30/2024 3:46 AM, Fred. Zwarts wrote:
>>>>>>>> Op 29.apr.2024 om 21:04 schreef olcott:
>>
>> [ .... ]
>>
>>>>> When we add the brand new idea of {simulating termination analyzer} to
>>>>> the existing idea of TM's then we must be careful how we define 
>>>>> halting
>>>>> otherwise every infinite loop will be construed as halting.
>>
>>
>>>> Why?
>>
>>>> That doesn't mean the machine reached a final state.
>>
>>
>>> Alan seems to believe that a final state is whatever state that an
>>> aborted simulation ends up in.
>>
>> Only through your twisted reasoning.  For your information, I hold to the
>> standard definition of final state, i.e. one which has no state following
>> it.  An aborted simulation is in some state, and that state is a final
>> one, since there is none following it.
>>
>>> On 4/30/2024 10:44 AM, Alan Mackenzie wrote:
>>>> You are thus mistaken in believing "abnormal" termination
>>>> isn't a final state.
>>
>>>> Only if you try to define something that is NOT related to Halting, do
>>>> you get into that issue.
>>
>>
>>> "The all new ideas are wrong" assessment.
>>> Simulating termination analyzers <are> related to halting.
>>
>> Except you cannot define what such a thing is, and that relationship is
>> anything but clear.
>>
> 
> When a simulating termination analyzer matches one of three
> non-halting behavior patterns
> (a) Simple Infinite loop
> (b) Simple Infinite Recursion
> (c) Simple Recursive Simulation

Except that (c) is NOT a correct non-halting pattern if the "Simulator" 
is a decider that may abort its simulation.

So, it isn't a "Non-Halting Pattern"

This was proven to you YEARS ago by showing a CORRECT simulation of D(D) 
calling the H(D,D) that used that definition.

So, unless you come clean that these are NOT patterns about "Halting" 
but something else, which needs a different name, you are just shown to 
be a LIAR.

> 
> It aborts it simulation and reports that the input specifies
> a non-halting sequence of configurations. Otherwise it continues
> to simulate the input to completion. Non-terminating inputs that
> have complex non-halting behaviors are outside of its domain.

Except that the input specifies a FINITE sequence of configurations, 
since the H that is calls WILL abort its simimulation of the input it is 
given and return 0.

YOUR logic is based on the LIE of CHANGING the H to be something else, 
because you just don't understand the nature of computer programs.

> 
> 
>>> The whole field of *termination analysis* is directly related
>>> to halting.
>>
>> Is there such a field of study?
>>
> 
> WST 2023: 19th International Workshop on Termination
> https://easychair.org/cfp/WST2023
> 
> https://en.wikipedia.org/wiki/Termination_analysis
> 
> *AProVE: Non-Termination Witnesses for C Programs*
> To prove (non-)termination of a C program, AProVE
> uses the Clang compiler [7] to translate it to the
> intermediate representation of the LLVM framework [15].
> Then AProVE symbolically executes the LLVM program ...
> https://verify.rwth-aachen.de/giesl/papers/TACAS22.pdf

Yes, there are a LOT of non-terminating programs that can be detectected.

The problem is that when you make H and D actual programs, if H(D,D) 
returns 0, then D(D) is NOT a "non-terminating" program.

Now, part of the issue is that this form of Termination Analysis isn't 
as concerned about being able to be 100% for every possible program, but 
wants to look at what classes of programs CAN be very reliably decided on.

So yes, it is RELATED to halting, but has a different criteria for what 
is considered a solution. In part, because they KNOW that 100% accuracy 
on EVERY program is impossible, so they want to study what CAN be done.

In the field, rejecting "hostile" programs that are trying to be 
intentionally hard to decide isn't considered a failure.


> 
>> [ .... ]
>>
>>> -- 
>>> Copyright 2024 Olcott "Talent hits a target no one else can hit; Genius
>>> hits a target no one else can see." Arthur Schopenhauer
>>
>