> It's certainly possible to write a program that always terminates with an answer of either HALTS or UNKNOWN, and only says HALT when that's true, it's just th

> It's certainly possible to write a program that always terminates with an answer of either HALTS or UNKNOWN, and only says HALT when that's true, it's just that it will also return UNKNOWN for some (or all) programs that do actually halt. Any program running in a Turing-complete environment is subject to the Halting Problem. So, given that constraint, your example program cannot be relied on to do any specific thing. That's the meaning of the Turing Halting Problem. https://en.wikipedia.org/wiki/Halting_problem https://en.wikipedia.org/wiki/Halting_problem : "Alan Turing proved in 1937 that the halting problem is undecidable, meaning that no general algorithm exists that can correctly solve the problem for all possible program–input pairs." Focus your attention on the word "undecidable".