Re: The philosophy of computation reformulates existing ideas on a new basis ---

Liste des GroupesRevenir à theory 
Sujet : Re: The philosophy of computation reformulates existing ideas on a new basis ---
De : polcott333 (at) *nospam* gmail.com (olcott)
Groupes : comp.theory
Date : 01. Nov 2024, 13:03:24
Autres entêtes
Organisation : A noiseless patient Spider
Message-ID : <vg2g2c$37lpn$4@dont-email.me>
References : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
User-Agent : Mozilla Thunderbird
On 11/1/2024 5:37 AM, Mikko wrote:
On 2024-10-31 23:43:41 +0000, olcott said:
 
On 10/31/2024 6:08 PM, Richard Damon wrote:
On 10/31/24 12:12 PM, olcott wrote:
On 10/31/2024 11:03 AM, Andy Walker wrote:
On 31/10/2024 11:01, Mikko wrote:
On 2024-10-30 11:17:45 +0000, Andy Walker said:
On 30/10/2024 03:50, Jeff Barnett wrote:
You may have noticed that the moron responded to your message in
less than 10 minutes. Do you think he read the material before
responding? A good troll would have waited a few hours before
answering.
    I doubt whether Peter is either a moron or a troll.
Does it really matter? If he falsely pretends to be a moron or a liar
I may politely pretend to believe.
>
     It's not exactly polite to describe Peter in any of these ways!
Entirely personally, I see no reason to do so in any case.  He is quite
often impolite in response to being called a "stupid liar" or similar,
but that's understandable.  He is no worse than many a student in terms
of what he comprehends;  his fault lies in [apparently] believing that he
has a unique insight.
>
When what I say is viewed within the perspective of
the philosophy of computation I do have new insight.
>
When what I say is viewed within the assumption that
the current received view of the theory of computation
is inherently infallible then what I say can only be
viewed as incorrect.
>
So, are you willing to state that you are admitting that nothing you might come up with has any bearing on the original halting problem because you are working in a new framework?
>
>
I am admitting one of two things:
(1) Everyone has misconstrued the original halting problem
as not applying to the behavior actually specified by the
actual input finite string.
 The finite strings specifying the behaviour are not a part
of the halting problem. Any solution is required to contain
encoding rules for the creation of those strings.
 
Sure they are. The halting problem is entirely about finite
string TM pairs when the finite string is a Turing machine
description that specifies its own behavior. If it was not
about the semantic property of the behavior specified by
this finite string then we are left with syntactic properties
that are known to be decidable.

(2) I am resolving the halting problem in a way that is
comparable to the way that ZFC resolved Russell's Paradox.
 Problems shall be solved, not resolved. The expression "resolving
the halting problem" does not mean anything because the types of
the words are not compatible. A paradox is a different type so
it can be resolved.
 
It is iffy to say that ZFC solved Russell's Paradox because
it is not solving the original problem it is redefining the
basis of the problem.

Establishing the foundation that the decider must report on
the behavior of its own simulation of its input to compute
the mapping from this input to its behavior.
 Establishing another foundation means that your work is about
something else than the halting problem.
No more so than ZFC is not about the naive set theory form
of Russell's Paradox.

Another foundation
may be useful in finding something but it cannot be a part of
any solution of the halting problem.
Useful in the exact same way that ZFC resolved Russell's
Paradox. Simply replace an incoherent basis with a coherent one.

 Every solution to a halting
problem is either a Turing machine and encoding rules or a proof
that no pair of a Turing machine and encoding rules is a solution.
 
You contradicted yourself there:
"a proof that no pair of a Turing machine and encoding
rules is a solution." *is not within the category of*
"Every solution to a halting problem"
--
Copyright 2024 Olcott "Talent hits a target no one else can hit; Genius
hits a target no one else can see." Arthur Schopenhauer

Date Sujet#  Auteur
3 Jul 25 o 

Haut de la page

Les messages affichés proviennent d'usenet.

NewsPortal