Path: news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail
From: Mikko <mikko.levanto@iki.fi>
Newsgroups: comp.theory
Subject: =?utf-8?Q?Re:_Can_Flibble=E2=80=99s_neos-based_solution_still_be_Turing_Complete=3F?=
Date: Tue, 27 May 2025 11:09:16 +0300
Organization: -
Lines: 69
Message-ID: <1013rvc$2hai3$1@dont-email.me>
References: <62aYP.724368$wBt6.719989@fx15.ams4> <100sqj1$ppn2$1@dont-email.me> <100uco8$182pt$1@dont-email.me> <100va8a$1d5lg$8@dont-email.me> <101186o$1tpgo$1@dont-email.me> <10121o0$22da5$5@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Tue, 27 May 2025 10:09:18 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="d45fc908e51ad5515b792556774cb03f";
	logging-data="2665027"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18oGCl+YyLu69POveA+v5od"
User-Agent: Unison/2.2
Cancel-Lock: sha1:lKNqjI+i7ZW8/ZJSxK3EtWi3Lvo=

On 2025-05-26 15:35:28 +0000, olcott said:

> On 5/26/2025 3:19 AM, Mikko wrote:
>> On 2025-05-25 14:42:17 +0000, olcott said:
>> 
>>> On 5/25/2025 1:18 AM, Mikko wrote:
>>>> On 2025-05-24 16:02:41 +0000, olcott said:
>>>> 
>>>>> On 5/23/2025 9:20 PM, Mr Flibble wrote:
>>>>>> Yes, **Flibble’s neos-based solution can still be Turing Complete as a
>>>>>> whole**, even though it **disallows programs from referencing the
>>>>>> decider**.
>>>>>> 
>>>>>> Let’s break this down precisely.
>>>>> 
>>>>> A more useful application of the term Turing Complete would be that
>>>>  ...
>>>> 
>>>> The only useful meaning is what the term actually means. Any other
>>>> meaning is harmful.
>>> 
>>> Analysis of complex theory of computation problems
>>> is much more effective at the higher levels of
>>> abstraction of higher level languages.
>>> 
>>> For example because the x86 language has relative
>>> addressing the underlying model of computation
>>> specified by the x86 language has unlimited memory
>>> thus is Turing complete.
>> 
>> The generic x86 language does not specify the mapping from addresses
>> to memory locations. Different x86 processors do it differently. A
>> particular processor may be able to shift the mapping of a part of
>> the address space to a previous or next block of a potentially
>> infinite memory. However, the usual models can pnly map it to a larger
>> finite memory.
>> 
>> None of which is irrelevan to my note that the only useful meaning is
>> what the term actually means.
>> 
>>> Turing complete cannot possibly make any actual
>>> difference at all as long as the model of computation
>>> has enough memory for the algorithm.
>> 
>> It doesn't as long as you can compute every function you want to compute.
>> But if you don't know what you will want then having a Turing complete
>> system is best you can have.
> 
> The best system is a system that actually exists.
> there are far too many errors of false assumptions
> in models that are only imagined to exist.

Assusmptions about a hypthetical system cannot be identified as false
because they cannot be compared to reality. It is possible that the
assumptions are found to be inconsisent, in which case we know that
some of the assumtions are false and that the system is not useful.
But we can prove various conseqneces of those assumptions and compare
them to our desiderata to determine whether a real world modes of the
system would be useful and whether it could be implermented in the
real world.

A systems that actually exist can have features that no one has thought
they could have, and as a consequence they may behave in ways that no
one understands. The best way to avoid that is to implement a well
understood theoretical system.

-- 
Mikko