Hacker News new | past | comments | ask | show | jobs | submit login

> Nope. You're thinking about this mathematically, not algorithmically.

I don't know what you mean by this, and I'm not sure how it relates to what I said. I'm using "decidable" in the strict, computer-science sense of the word.

The statement in example 1 of the paper, which we're discussing, is about computable real numbers. A computable real number is one for which there is a Turing machine that can calculate it to any desired finite precision in finite time.

A semidecidable problem is one for which there is a program that halts if the answer is "yes", but may not halt if the answer is "no". The halting problem is in this category.

Given a computable real number x, asking whether x<0 or x>0 are semidecidable but not decidable problems.




Yes, you're right. See https://news.ycombinator.com/item?id=41154273 for my post-mortem.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: