Re: Two computer science professors agree with Flibble

On 5/6/25 1:37 PM, olcott wrote:

On 5/6/2025 6:25 AM, Richard Damon wrote:

On 5/5/25 9:31 PM, olcott wrote:

On 5/5/2025 8:16 PM, Richard Damon wrote:

On 5/4/25 11:10 PM, olcott wrote:

On 5/4/2025 10:00 PM, dbush wrote:

On 5/4/2025 9:38 PM, olcott wrote:

On 5/4/2025 8:13 PM, Ben Bacarisse wrote:

Richard Heathfield <rjh@cpax.org.uk> writes:
>

On 04/05/2025 23:34, Mr Flibble wrote:

The function is neither computable nor incomputable because there is no
function at all, just a category error.

>
It's a point of view.

>
It's a point of view only in the sense that there is no opinion so daft
that it's not someone's point of view.  The technical-sounding waffle
about it being a "category error" is simply addressed by asking where
the supposed category error is in other perfectly straightforward
undecidable problems.  For example, whether or not a context-free
grammar is ambiguous or not, or the very simple to pose Post
correspondence problem.
>

>
Flibble IS CORRECT when the halting problem is defined
to be isomorphic (AKA analogous) to the Liar Paradox:
"This sentence is not true".
>
When the Halting Problem is defined as an input that
does the opposite of whatever its decider reports
then both Boolean return values are incorrect

>
False.  One value is correct and one is incorrect.
>

>
Both Boolean RETURN VALUES FROM H *ARE* INCORRECT,
Even though D halts or fails to halt.

>
No, the given H can only return one of the values.
>
The other one is correct.
>

>
The linguistic context of WHO IS ASKED is an essential
part of the question.

>
No, because H^ has the same behavior to all deciders, it only makes H wrong, as it behaves the opposite of whichever is the one answer that H gives.
>

>
Math and Comp Sci people that are clueless about these
details of how language actually works think that they
can get away with ignoring a crucial part of the actual
question.

>
No, you are clueless as to the requirements of H being a program / fixed algorithm.
>

>
That is NOT what professor Sipser agreed to.
>
<MIT Professor Sipser agreed to ONLY these verbatim words 10/13/2022>
     If simulating halt decider H correctly simulates its input D
     until H correctly determines that its simulated D would never
     stop running unless aborted then
>
*would never stop running unless aborted*
is one actual input and the hypothetical
HHH/DD that never aborts.
>

>
But hypothetical DD isn't the DD that was given to the original HHH,

 Yet <is> the HHH/DD that professor Sipser agreed to:
*would never stop running unless aborted*
*would never stop running unless aborted*
*would never stop running unless aborted*
 The above refers to an HHH that does not abort.
The above refers to an HHH that does not abort.
The above refers to an HHH that does not abort.
 

No, H *IS* the one and only decider actually mentioned.
D *is* the one and only program do decide on which calls that H.
This H is allowed to abort its simulation, if it can prove that a CORRECT SIMULATION of that input will not reach a final state.
If you have your H use that option, then the H that aborts is the H mentioned everywhere and the H that the D calls.
Since that H doesn't do a correct simulation, you can't use what it does to prove the result, but can give that exact input to a hypothetical H that doesn't abort.
When we look at that, because the input is still the input it always was, and thus still calls the H that aborts, we find that a correct simulation of that input will halt.
Thus, the H that tried to use the rule to abort, finds that it didn't actually have the right to use it, so it made an error,
It can't "go back in time" to remove that decision, as that is not an allowed finite string transformation.
Your attempt at unsound logic just shows your utter ignorance of the fields you are attempting to talk about.
You don't seem to understand that a program is what it is, and is fully defined, and thus we can't make D until you fix H, and then D if built fixed to THAT H, and then when you ask about changing H to something different, it doesn't go back in time to change those earlier decisions.
The problem is you just can't build a D that does what you want, and changes itself based on the decider that is deciding it. Your attempts just prove you don't know what you are talking about.