Comment by ricardobayes
3 years ago
Ever heard of real-time systems? edit: man this place is filling up with pseudo-intelligent naysayers at an alarming pace.
3 years ago
Ever heard of real-time systems? edit: man this place is filling up with pseudo-intelligent naysayers at an alarming pace.
How can a system solve the halting problem?
The halting problem states that it's impossible to prove if any arbitrary program halts. This has no relevance to latency of real software. Software is laggy because programmers pile up abstractions without caring about latency, not because of an abstract mathematical theorem. It's possible to write hard real-time software with provable latency bounds.
It is possible, but the proof's effort grows combinatorially with the size of the program. The large majority of "real software" is not amenable to this kind of test, making the mathematical theorem very concrete.
4 replies →
You don’t have to solve the halting problem to have consistent latency - you are conflating distinct concepts. The fact that a computer can’t check if an arbitrary program will eventually halt does not preclude the existence of programs that halt with exactly the same number of instructions every single time.