Re: Can D simulated by H terminate normally?

On 2024-05-01 19:44, olcott wrote:

On 5/1/2024 7:44 PM, André G. Isaak wrote:

On 2024-05-01 18:16, Richard Damon wrote:

On 5/1/24 11:26 AM, olcott wrote:

>

It is a termination analyzer thus is not required to be infallibly
correct on every possible input. It must get at least one input
correctly.

>
In other words, it is a TOY.
>
By your definition:
>
H(ptr m, ptr d) {
     return 1;
}
>
is a correct termination analyzer, as it will get at least one input correctly.

>
Actually, by the metric which he gives, every single decider in existence is a correct termination analyzer, which tends to suggest this metric is relatively useless.
>
André
>

 Try and back that up with reasoning anchored in quotes from my paper.
 https://www.researchgate.net/publication/369971402_Termination_Analyzer_H_is_Not_Fooled_by_Pathological_Input_D

I was commenting on the line which you wrote here in this group, not a line from the above 'paper'.
Since the above paper talks about "termination analyzers" without offering any definition of what is meant by this term, I can only go by your comment in this group where you write that "a termination analyzer thus is not required to be infallibly correct on every possible input. It must get *at least one* input correctly." [emphasis mine].
That means that any decider which correctly accepts *one* terminating program as terminating is a "termination analyzer" (unless, of course, you mean something entirely different by "termination analyzer". But this would require you to supply an actual definition.)
And every decider will accept at least *one* terminating program description, regardless of what it is that that decider was actually intended to decide. That follows from the simple fact that it is no more possible to construct a decider which gets every instance of the halting problem* wrong than it is to construct one that gets every instance right.
André
[*] Here talk about 'halting' as if it is the same thing as 'terminating'. Since you've switched from talking about halt deciders to 'termination analyzers', perhaps you think these mean different things. But that again would require you to actually define what you think this difference is. Otherwise we're left simply trying to guess the meanings of your terms, as usual.
--
To email remove 'invalid' & replace 'gm' with well known Google mail service.