XPost: comp.theory, sci.logic
On 7/26/25 1:59 PM, olcott wrote:
On 7/26/2025 12:31 PM, Alan Mackenzie wrote:
Hello, Ben.
Ben Bacarisse <[email protected]> wrote:
Alan Mackenzie <[email protected]> writes:
[ .... ]
In comp.theory olcott <[email protected]> wrote:
On 7/21/2025 10:52 AM, Alan Mackenzie wrote:
...
More seriously, you told Ben Bacarisse on this newsgroup that you had >>>>>> fully worked out turing machines which broke a proof of the Halting >>>>>> Theorem. It transpired you were lying.
Just for the record, here is what PO said late 2018 early 2019:
On 12/14/2018 5:27 PM, peteolcott wrote that he had
"encoded all of the exact TMD [Turing Machine Description]
instructions of the Linz Turing machine H that correctly decides
halting for its fully encoded input pair: (Ĥ, Ĥ)."
Date: Sat, 15 Dec 2018 11:03:21 -0600
"Everyone has claimed that H on input pair (Ĥ, Ĥ) meeting the Linz >>> specs does not exist. I now have a fully encoded pair of Turing
Machines H / Ĥ proving them wrong."
Date: Sat, 15 Dec 2018 01:28:22 -0600
"I now have an actual H that decides actual halting for an actual (Ĥ, >>> Ĥ) input pair. I have to write the UTM to execute this code, that >>> should not take very long. The key thing is the H and Ĥ are 100% >>> fully encoded as actual Turing machines."
Date: Sun, 16 Dec 2018 09:02:50 -0600
"I am waiting to encode the UTM in C++ so that I can actually execute >>> H on the input pair: (Ĥ, Ĥ). This should take a week or two [...] it >>> is exactly and precisely the Peter Linz H and Ĥ, with H actually
deciding input pair: (Ĥ, Ĥ)"
Date: Fri, 11 Jan 2019 16:24:36 -0600
"I provide the exact ⊢* wildcard states after the Linz H.q0 and after
Ĥ.qx (Linz incorrectly uses q0 twice) showing exactly how the actual >>> Linz H would correctly decide the actual Linz (Ĥ, Ĥ)."
Thanks for clarifying that.
I think I can understand a bit what it must feel like to be on the
receiving end of all this. Firstly you know through training that what
you're being told is definitely false, but on the other hand you don't
like to believe that somebody is lying; somehow you give them the
(temporary) benefit of the doubt. Then comes the depressing restoration
of truth and reality.
When the topic came up again for
discussion, you failed to deny writing the original lie.
That is the closest thing to a lie that I ever said.
When I said this I was actually meaning that I had
fully operational C code that is equivalent to a
Turing Machine.
I think it was a full blown lie intended to deceive. Did you ever
apologise to Ben for leading him up the garden path like that?
No, never. In fact he kept insulting me until it became so egregious
that I decided to having nothing more to do with him.
Somehow, that doesn't surprise me. I only post a little on this group
now (I never really posted much more) for similar reasons. I care about
the truth, including mathematical truth; although I've never specialised
in computation theory or mathematical logic, I care when these are
falsified by ignorant posters.
What really got my goat this time around was PO stridently and
hypocritically accusing others of being liars, given his own record.
What he did do was take months to slowly walk back the claim he made in
December 2018. H and Ĥ became "virtual machines" and then started to be >>> "sufficiently equivalent" to Linz's H and Ĥ rather the "exactly and
precisely the Peter Linz H and Ĥ". By Sep 2020 he didn't even have it >>> anymore:
"I will soon have a partial halt decider sufficiently equivalent to >>> the Linz H correctly deciding halting on the Linz Ĥ"
It took nearly two years to walk back the clear and explicit claim to
this vague and ill-defined claim of not having something!
Yes. I've watched the latter part of this process.
You have not and never have had "fully operational C code" that
breaks a
proof of the Halting Theorem. To say you had this, when you clearly
didn't, was a lie.
He also tried to pretend that the C code (which, as you say, he didn't
have) is what he always meant when he wrote the words I quoted above. I >>> defy anyone to read those words with PO's later claim that he meant C
code all along and not conclude that he was just lying again to try to
save some little face.
What amazes me is he somehow thinks that theorems don't apply to him.
Of course, he doesn't understand what a theorem is, somehow construing
it as somebody's opinion. If it's just opinion, then his contrasting
opinion must be "just as good". Or something like that.
C code does not have "TMD instructions" that can be encoded. TMs (as in >>> Linz) do. When executed, C code has no "exact ⊢* wildcard states after >>> the Linz H.q0" for PO to show. A TM would. C code does not need a UTM >>> to execute it (a TM does) and if he really meant that he had C code all
along, does anyone think he could write a UTM for C in "a week or two"?
It is so patently obvious that he just had a manic episode in Dec 2018
that caused he to post all those exuberant claims, and so patently
obvious that he simply can't admit being wrong about anything that I
ended up feeling rather sorry for him -- until the insults started up
again.
That's another reason I don't post much, here. I really don't feel like
being insulted by somebody of PO's intellectual stature.
Have a good Sunday!
--
Ben.
The error of all of the halting problem proofs is
that they require a Turing machine halt decider to
report on the behavior of a directly executed
Turing machine.
Because that *IS* the definition of the problem.
It is common knowledge that no Turing machine decider
can take another directly executing Turing machine as
an input, thus the above requirement is not precisely
correct.
And where is that "common knowledge" comming from. Since we CAN
represent a Turing Machine with enough detail for a UTM to recreate the behavior, the requirement is valid.
When we correct the error of this incorrect requirement
it becomes a Turing machine decider indirectly reports
on the behavior of a directly executing Turing machine
through the proxy of a finite string description of this
machine.
But the only "error" is you LIE that a Turing Machine can report on
behavior of the machine properly represented to it.
ALl you are doing is proving you are just a stupid pathological liar.
The fact that you are don't understand how to do it, doesn't mean it
can't be done. It just proves your stupidity, since it has been proven
that it can be done.
What can't be done is to determine that behavior that requires an
unbounded number of steps to be shown, can necessarily be determined in
a bounded number of steps.
Now I have proven and corrected the error of all of the
halting problem proofs.
No, you have proven that you have no idea what is correct and what is
false, and thus turned yourself into a pathological liar, a liar that
has lost the ability to tell the differene between truth and lies.
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)