Re: Alan Turing's Halting Problem is incorrectly formed (PART-TWO)

From: Bryan Olson (fakeaddress_at_nowhere.org)
Date: 06/29/04 

Date: Tue, 29 Jun 2004 02:25:59 GMT

Peter Olcott wrote:

> "Bryan Olson" wrote:
>
>>Peter Olcott wrote:
>>
>> >> > bool LoopIfHalts (bool YouSayItHalts)
>> >> > {
>> >> > if (YouSayItHalts)
>> >> > while(true)
>> >> > ;
>> >> > else
>> >> > return false;
>> >> > }
>> >> >
>> >> > This function would both compile and run correctly.
>> >>
>> >>And I can describe another program that decides whether
>> >>LoopIfHalts(x) halts for any given x. Working at that level is
>> >>not going to lead you to an understanding of Turing's result.
>> >
>> > NO YOU CAN'T That's the whole point.
>>
>>LoopIfHalts halts given 'false', runs forever given 'true', and
>>essentially hangs for non-boolean inputs. In standard C++:
>>
>> bool LoopIfHalts_halts_for_input(bool input)
>> {
>> return ! input;
>> }
>
>
> No when you provide "true" to the LoopIfHalts
> function you are also saying "This program halts"

No, check your references. I'm not also saying that, and the
halting problem has nothing to do with whether I'm saying that.
As the halting problem is defined, I've produced a machine that
solves the special case of the halting problem where the machine
is LoopIfHalts.

If you want to re-define the same terms to refer to some other
problem, I'm not really interested. Especially if all you can
argue is that your version doesn't make sense.

> That is why I called it YouSayItHalts. As soon as
> You Say It Halts is goes into an infinite loop.

But that's not what I said. I said it "halts given 'false',
runs forever given 'true', and essentially hangs for non-
boolean inputs". I was right. I said I could "describe another
program that decides whether LoopIfHalts(x) halts for any given
x." You said I could not, and you were proven wrong. If you
have a counter-example, let me know, but don't make up stories
that I also said something I didn't. 

-- 
--Bryan

Relevant Pages
Loading