On 5/7/2021 8:25 AM, Ben Bacarisse wrote:
> olcott <No...@NoWhere.com> writes:
>
>> On 5/6/2021 7:47 PM, Ben Bacarisse wrote:
>>> olcott <No...@NoWhere.com> writes:
>>>
>>>> On 5/6/2021 6:39 PM, Ben Bacarisse wrote:
>>>>> olcott <No...@NoWhere.com> writes:
>>>>>
>>>>>> On 5/6/2021 5:53 PM, Ben Bacarisse wrote:
>>>>>>> olcott <No...@NoWhere.com> writes:
>>>>>>>
>>>>>>>> On 5/6/2021 4:22 PM, Ben Bacarisse wrote:
>>>>>>>>> olcott <No...@NoWhere.com> writes:
>>>>>>>>>
>>>>>>>>>> On 5/6/2021 10:26 AM, Ben Bacarisse wrote:
>>>>>>>>>>> olcott <No...@NoWhere.com> writes:
>>>>>>>>>>>
>>>>>>>>>>>> Linz H decides not halting on Linz Ĥ on the basis of infinite recursion.
>>>>>>>>>>>>
http://www.liarparadox.org/Peter_Linz_HP(Pages_315-320).pdf
>>>>>>>>>>>
>>>>>>>>>>> No. In order to save time (yours and mine), I've said I'll ask if you'd
>>>>>>>>>>> like to learn this material before taking the time to explain. Do say
>>>>>>>>>>> if you have time to learn why this is wrong.
>>>>>>>>>>
>>>>>>>>>> More literally for every at least partial halt decider H that bases
>>>>>>>>>> its halting decision on simulating its input unless H decides
>>>>>>>>>> infinitely nested simulation (not halting), stops simulating Ĥ(Ĥ), and
>>>>>>>>>> decides not halting the alternative is that H never returns and Ĥ
>>>>>>>>>> remains in infinitely nested simulation.
>>>>>>>>>
>>>>>>>>> Well that saves me some time. You don't want to know why you were wrong
>>>>>>>>> in the previous post.
>>>>>>>>
>>>>>>>> You can say that I am wrong and this has as much actual truth behind
>>>>>>>> is as Trump won the election by a unanimous consensus.
>>>>>>>>
>>>>>>>> Any dishonest lout can falsely claim that someone is wrong.
>>>>>>>
>>>>>>> Yes. And someone knowledgeable a truthful like me could tell you what
>>>>>>> you got wrong, but you told me you don't have time for that, so I won't
>>>>>>> unless you ask me to. And I'm only offering to do that for the first
>>>>>>> mistake above.
>>>>>>
>>>>>> _Infinite_Loop()...
>>>>>
>>>>> I am offering to explain one thing: why "Linz H decides not halting on
>>>>> Linz Ĥ on the basis of infinite recursion" is wrong but you keep
>>>>> ignoring this offer. Either ask me to explain (and take the trouble to
>>>>> read what then write) or leave the thread alone.
>>>>>
>>>>
>>>> Try and refute this more accurately stated basis:
>>>>
>>>> Ĥ.q0 wM ⊢* Ĥ.qx wM wM ⊢* Ĥ.qy ∞
>>>> Ĥ.q0 wM ⊢* Ĥ.qx wM wM ⊢* Ĥ.qn
>>>>
>>>> The above is adapted from (Linz:1990:319).
>>>> It shows that Turing machine Ĥ copies its input at (q0) and begins
>>>> executing an embedded copy of the original halt decider with this
>>>> input at (qx).
>>>
>>> I like your use of . to identify the TM a state belongs to, but you
>>> never use it to its full potential. Instead of having to say what qx
>>> is, you could write H.q0.
>>
>> That would be incorrect.
>> The code from H that is now embedded within Ĥ is no longer H.
>
> You can call the states of H^ anything you like. Using H.s for those
> copied from H just helps talk about them
>
>>> And you could show the H'.qa and H'.qb states
>>> rather the describing them in words.
>>
>> It is no longer H' states. The whole thing is now a single contiguous
>> Ĥ
>
> Then there is no point in using the dot. It just gets in the way.
>
>>>> It can be understood from the above specification that when the
>>>> embedded halt decider @Ĥ.qx bases its halting decision on simulating
>>>> its input, and it has (Ĥ, Ĥ) as its input that:
>>>
>>>> Ĥ.q0 would copy its input and then Ĥ.qx would simulate its input with this copy then
>>>> Ĥ.q0 would copy its input and then Ĥ.qx would simulate its input with this copy then
>>>> Ĥ.q0 would copy its input and then Ĥ.qx would simulate its input with this copy...
>>>> unless and until the halt decider @Ĥ.qx stops simulating its input.
>>>
>>> Yes. I wonder why you think this is in doubt.
>>
>> It means that the halting problem instance is decidable as not
>> halting.
>
> No. Explanation available on request.
>
>>>> Linz, Peter 1990. An Introduction to Formal Languages and
>>>> Automata. Lexington/Toronto: D. C. Heath and Company.
>>>
>>> Take care here. This makes it look like Linz wrote the silly stuff
>>> about simulation. He didn't.
>>>
>>> Does this mean you /would/ like me to say why "Linz H decides not halting
>>> on Linz Ĥ on the basis of infinite recursion" is wrong? Maybe you can
>>> see it for yourself now?
>>
>> As I said infinite recursion is not precisely correct it is actually
>> infinitely nested simulation.
>
> So you don't want your original mistake explained? OK. Why not just
> say "no", rather than blather?
>
>> You just agreed that for every at least partial halt decider H that is
>> based on simulating its input that Ĥ(Ĥ) does specify infinitely nested
>> simulation unless H stops simulating it.
>
> The nested simulations are finite in number. What /would/ happen unless
> something or other were done is irrelevant. What matters is what /is/
> done, and the nested simulations are stopped and are therefore not
> infinite.
>
>> So what is left that you could possibly say is incorrect:
>
> The same as ever: that Halts(H_Hat, H_Hat) is false when the arguments
> denote a finite computation. I.e. that it gives the wrong answer. You
> can't get round this with words.
>
void Infinite_Loop()
{
HERE: goto HERE;
}
int main()
{
Halts(Infinite_Loop(), Infinite_Loop());
}
How is it that you are not making the mistake of calling Infinite_Loop()
a finite computation because its simulation was stopped?
That would be like called a dead cow alive on the basis that it moves
when the rendering plant workers pick up its corpse.
> Also incorrect is that you reject one or other or both of these:
>
> (A1) For every instance of the halting problem that represents a finite
> computation, the correct answer is "yes".
> (A2) For every instance of the halting problem that does not represent
> a finite computation, the correct answer is "no".
>
> You keep trying to make the wrong answer sound reasonable, but even if
> you could, you are not saying anything about the halting problem because
> I tell /you/ what the right answers are for that problem.
>
If you are saying that infinite loops are not infinite then what you are
saying makes much less sense than what I am saying and others will know
this and agree.
> Your not-the-halting-problem is, none the less, interesting because, in
> general, it too is undecidable, but you don't know that yet.