Path: ...!feed.opticnetworks.net!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Mikko <mikko.levanto@iki.fi>
Newsgroups: comp.theory
Subject: Re: H(D,D) cannot even be asked about the behavior of D(D)
Date: Mon, 17 Jun 2024 10:10:02 +0300
Organization: -
Lines: 72
Message-ID: <v4onga$hjo3$3@dont-email.me>
References: <v45tec$4q15$1@dont-email.me> <v4cjl7$1o4b4$1@dont-email.me> <v4d991$3qbnc$1@i2pn2.org> <v4da12$1sioe$1@dont-email.me> <v4dbmf$3qbnc$3@i2pn2.org> <v4dcd6$1sioe$3@dont-email.me> <v4df0h$3qbnd$1@i2pn2.org> <v4dhf5$1tsdf$2@dont-email.me> <v4dja1$3qbnd$5@i2pn2.org> <v4djhf$1tsdf$6@dont-email.me> <v4dk7b$3qbnc$8@i2pn2.org> <v4dl3b$225kb$1@dont-email.me> <v4dn5u$3qbnd$8@i2pn2.org> <v4dop4$22o4a$2@dont-email.me> <v4dq07$3qbnc$12@i2pn2.org> <v4dqq0$2353n$1@dont-email.me> <v4el9m$3rsd6$3@i2pn2.org> <v4f3ec$2akmh$2@dont-email.me> <v4g65a$3tn6q$1@i2pn2.org> <v4g6vr$2ic0g$1@dont-email.me> <v4gc0b$3tn6r$6@i2pn2.org> <v4gcjc$2msea$1@dont-email.me> <v4geab$3tn6r$8@i2pn2.org> <v4gg0s$2nim8$2@dont-email.me> <v4ha63$3v16r$2@i2pn2.org> <v4hfq9$2sdqr$5@dont-email.me> <v4hp3r$3viml$1@i2pn2.org> <v4hv85$3021v$1@dont-email.me> <v4ju8f$222a$1@i2pn2.org> <v4k1m4$3f99u$1@dont-email.me> <v4k4mt$3fnqu$1@dont-email.me> <v4maeo$3vv3f$1@dont-email.me> <v4mnim$1qt6$6@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 17 Jun 2024 09:10:02 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="4dd19a4a6eb99e2a49c355dca3cdd6b7";
	logging-data="577283"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/wZX2M/qNJ2WrPLxCiMpFx"
User-Agent: Unison/2.2
Cancel-Lock: sha1:RTWpZiMiHTdcImKylAr2u4b0mBg=
Bytes: 4930

On 2024-06-16 12:59:02 +0000, olcott said:

> On 6/16/2024 4:15 AM, Mikko wrote:
>> On 2024-06-15 13:24:45 +0000, olcott said:
>> 
>>> On 6/15/2024 7:33 AM, Mikko wrote:
>>>> On 2024-06-15 11:34:39 +0000, joes said:
>>>> 
>>>>> Am Fri, 14 Jun 2024 12:39:15 -0500 schrieb olcott:
>>>>>> On 6/14/2024 10:54 AM, joes wrote:
>>>>>>> Am Fri, 14 Jun 2024 08:15:52 -0500 schrieb olcott:
>>>>>>>> On 6/14/2024 6:39 AM, Richard Damon wrote:
>>>>>>>>> On 6/14/24 12:13 AM, olcott wrote:
>>>>>>>>>> On 6/13/2024 10:44 PM, Richard Damon wrote:
>>>>>>>>>>> On 6/13/24 11:14 PM, olcott wrote:
>>>>>>>>>>>> On 6/13/2024 10:04 PM, Richard Damon wrote:
>>>>>>>>>>>>> On 6/13/24 9:39 PM, olcott wrote:
>>>>>>>>>>>>>> On 6/13/2024 8:24 PM, Richard Damon wrote:
>>>>>>>>>>>>>>> On 6/13/24 11:32 AM, olcott wrote:
>>>>> 
>>>>>>>> When H and D have a pathological relationship to each other then
>>>>>>>> H(D,D) is not being asked about the behavior of D(D). H1(D,D) has no
>>>>>>>> such pathological relationship thus D correctly simulated by H1 is the
>>>>>>>> behavior of D(D).
>>>>> What is H1 asked?
>>>>>>> H is asked whether its input halts, and by definition should give the
>>>>>>> (right) answer for every input.
>>>>>> If we used that definition of decider then no human ever decided
>>>>>> anything because every human has made at least one mistake.
>>>>> Yes. Humans are not machines.
>>>>>> I use the term "termination analyzer" as a close fit. The term partial
>>>>>> halt decider is more accurate yet confuses most people.
>>>> 
>>>> Olcott has used the term "termination analyzer", though whether he knows
>>>> what it means is unclear.
>>>> 
>>> 
>>> 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 
>>> and uses abstraction to obtain a finite symbolic execution graph (SEG) 
>>> containing all possible program runs.
>> 
>> AProVE is a particular attempt, not a defintion.
>> 
> 
> If you say: What is a duck? and I point to a duck that
> *is* what a duck is.

That would be just an example, not a definition. In particular, it does
not tell about another being whether it can be called a "duck".

> *Termination analysis*
> In computer science, termination analysis is program analysis which
> attempts to determine whether the evaluation of a given program halts
> for each input. This means to determine whether the input program
> computes a total function.
> https://en.wikipedia.org/wiki/Termination_analysis
> 
> I pointed out AProVE because it is essentially a simulating
> halt decider with a limited domain.

A difference between AProVE and a partial halt decider is that the input
to AProVE is only a program but not an input to that program but the
input to a partial halt decider contains both.

>>> *AProVE: Non-Termination Witnesses for C Programs*
>>> https://link.springer.com/content/pdf/10.1007/978-3-030-99527-0_21.pdf

-- 
Mikko