Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.software-eng > #2654

Re: A UTM halt decider <is> a pure function of its inputs (HP undecidability)

From Kaz Kylheku <563-365-8930@kylheku.com>
Newsgroups comp.theory, comp.lang.c, comp.lang.c++, comp.software-eng
Subject Re: A UTM halt decider <is> a pure function of its inputs (HP undecidability)
Followup-To comp.theory
Date 2021-03-30 15:03 +0000
Organization A noiseless patient Spider
Message-ID <20210330075406.357@kylheku.com> (permalink)
References (19 earlier) <hqidnTrQ-oVUE__9nZ2dnUU7-eudnZ2d@giganews.com> <20210329202056.955@kylheku.com> <uc-dnTZHvsKqPf_9nZ2dnUU7-Y_NnZ2d@giganews.com> <20210329211550.916@kylheku.com> <oLydnQPuAMc8rv79nZ2dnUU7-VPNnZ2d@giganews.com>

Cross-posted to 4 groups.

Followups directed to: comp.theory

Show all headers | View raw


On 2021-03-30, olcott <NoOne@NoWhere.com> wrote:
> On 3/30/2021 12:28 AM, Kaz Kylheku wrote:
>> On 2021-03-30, olcott <NoOne@NoWhere.com> wrote:
>>> Without changing the definition of H^ in the tiniest little bit I have
>>> shown that it specifies infinite recursion:
>> 
>> That is untrue.
>> 
>> What H^ specifies depends on the nature of H, because the bulk of H^ is
>> the sub-function H.
>
> http://www.liarparadox.org/Peter_Linz_HP(Pages_318-319).pdf
>
> Whenever H is a simulator then H^ <is> infinitely recursive never 
> reaching the appended loop states whether or not the simulator always 
> returns a value of true or false.

H is not required to be a simulator.

If H simply executes its input, we know that it fails to decide due to
runaway recursion: the input calls the decider, which executes its
input, which calls the decider.

This is freshman-level obvious.

The first materials I ever read about the halting problem spelled this
out, for the dummies.

> Never has had any effect on halting problem undecidability

Not returning is completely, utterly relevant to undecidability.

The undecidability of halting means that a decision function can
only decide at most a subset of the possible cases, not all of them.

Outside of the subset which it decides, it misbehaves in one of several
ways: it either returns the wrong answer, or fails to halt, or else
something else, like failing due to a division by zero or similar
problem.

Runaway recursion falls under "fails to halt"; it's a failed case.

> because no halt decider would ever return any value to H^.

Except in your program, where a "pure function" sometimes returns a
value and sometimes doesn't, for the same pair of visible arguments!

-- 
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal

Back to comp.software-eng | Previous | Next — Previous in thread | Find similar | Unroll thread


Thread

Re: A UTM halt decider <is> a pure function of its inputs (HP undecidability) olcott <NoOne@NoWhere.com> - 2021-03-30 09:20 -0500
  Re: A UTM halt decider <is> a pure function of its inputs (HP undecidability) Kaz Kylheku <563-365-8930@kylheku.com> - 2021-03-30 15:03 +0000

csiph-web