From Newsgroup: comp.ai.philosophy

On 3/26/2026 1:16 AM, Chris M. Thomasson wrote:

On 3/25/2026 6:13 PM, Richard Damon wrote:
[...]

I tried with one that will never halt until all branches have been hit,
in some fuzzer code, even in AppleSoft BASIC, lol. The fuzzer is useful
in many things. Especially tricky lock-free programming...

A song for dart:

https://youtu.be/ko70cExuzZM?list=RDSdq4T3iRV80

rofl!
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.ai.philosophy

On 3/25/26 10:07 PM, dart200 wrote:

On 3/25/26 6:13 PM, Richard Damon wrote:

On 3/25/26 8:09 PM, dart200 wrote:

On 3/25/26 3:49 PM, Richard Damon wrote:

On 3/25/26 12:26 PM, dart200 wrote:

On 3/25/26 4:02 AM, Richard Damon wrote:

how are you so sure there even exits a machine which is so
undecidable that no machine can decide on it?

That is a different problem, and not needed for this one.

u literaly said:

it isn't UNDECIDABLE, because there exists machines that
can decide on its behavior.

so what is that UNDECIDABLE machine where classification fails
for _all_ classifiers???

We can't know what it is.

It seems you are too stupid to understand that concept.

It can be proven that there exists some machines that run forever >>>>>> that no known to be always correct machine can determine this
fact, because there is not detectable pattern in its operation.

Note, even those machines aren't actually called "undecidable", as >>>>>> "undeciable" is a property of a problem, not the instance of the
problem on a given machine, and it turns out we can't name a
machine that we know we can not determine its halting property.

All we can do is build up a class of machines we can't determine
the answer for yet, with the knowledge that in the fullness of
that set, there are machies that we will never be able to know
what they do.

Some machines will leave that set as we do push on them long
enough to classify, as the work level needed to classify them was >>>>>> reached, but for some machines, that work level is actually
infinite, so we can never reach it.

it hasn't been proven that there exists a machine which cannot be
understood,

it's only been proven that a single TM can't output all the answers... >>>>>

No, there is a proof that there is a machine that we can not know if
it halts or not, that NO decider, even a partial but always correct
decider, can tell us that it doesn't halt. (We of course can
eventually determine that any machibe that halts will halt, so the
unknowable machines must be non-haltiing).

The proof is just too abstract for you to understand, and I would
need to find it again to show you.

ok so do that eh???

WhY?

Have you made the first step and decided to agree that the base
halting problem has been proved uncomputable?

i have agreed that a naive halting decider can't be implemented as
specified by turing (and others), and i have agreed with that many a
times now

Ig isn't just a "naive" halt decider, it is about ANY Halt Decider that
wants to try to claim to be always correct.

The "Pathological" / "Self-Referencial" machine built from that decider
*WILL* be decided incorrectly by that decider, so it just fails to meet
the requirement.

Thus *NO* Always Correct Halt Decider can exist.

It is YOUR "naive" idea that you can change the requirements that is wrong.

Just like Turing's proof that you can't compute the diagonal of ALL circle-free machine outputs, which implies that you can not compute an enumeration of all circle-free machines.

And, a similar proof that he makes mention of shows that similarly, you
can not compute the enumeration of computable numbers, as that would
enable you to build a machine that computes the anti-diagonal, meaning
the anti-diagonal is a computable number, but won't be in the
enumeration that was supposed to contain ALL computable numbers, and
thus that enumeration can't be computed (even though such an enumeration exists by the counting principle).

Thus, you can't "fix" the Diagonal, as your "fixed diagonal" isn't the diagonal required in the problem, but is of just a partial enumeration,
not the complete enumeration.

If not, you won't beleive the more abstract one at all.

instead of gaslighting me more u could just post the proof ur thinking of

I would need to find the article that mentioned it, and this was years
ago, and might not be in something that is free online.

Since you have proved your stupidity, and wouldn't understand it anyway,
it isn't worth the effort.

Since you don't even understand what a single machine is, that topic
is WELL beyond your ability.

are you expecting me to accept something without understanding it...

No, I don't/

or you don't,

and ur just telling me to soothe your ego?

I mentioned it as an aside that shows how deep uncomputability goes.

When you make the claim of the opposite, I point out that you HAVE
been proved wrong, but are too ignorant.

Since you have shown that you don't understand the meaning of the
basic words, it doesn't make sense to try to go deeper.

Let me ask you, where did you get your WRONG definitions of the core
words, as they are wrong?

What do YOU think a "Computation" would be?

What do YOU think a "Algorithm" consists of?

What do you think "Computability" means?

What do you think a "Program" is?

and where, other than your ignorant head, did you get those ideas from?

The problem is, these words HAVE actual definitions agreed by the
community, that define the field, and to use diffferent defintions
means you are not actually working in the field, but in something
else, and thus LIE when you claim you are working in the field.

--- Synchronet 3.21f-Linux NewsLink 1.2