Skip to content

Comment on Did Turing prove the undecidability of the halting problem?parent

Comments

Formal verification (i.e., automated theorem proving) relies on only admitting 'total' programs, which usually requires that any potential sources of 'bottom' (i.e., non-termination) come equipped with a proof that that never happens (usually based on well ordering of some structural decomposition in the recursion).

Thus, most theorem provers are not turing complete and explicitly ban any turing complete structures.

So, the various things they show and prove cannot be said to encompass all programs.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.