Re:Proving false in Coq using an implementation bug
by seanmcdirmid on Tuesday March 24, 2015 @07:40PM
(#9260342)
You can re-run your proofs once the bug is patched, so its not like this makes all those proofs useless. Actually, that is the nice thing about Coq.
[ Parent
| on HN ]
Re:Proving false in Coq using an implementation bug
by jsprogrammer on Tuesday March 24, 2015 @08:05PM
(#9260460)
How do you know there isn't another, yet undiscovered, bug?
Re:Proving false in Coq using an implementation bug
by seanmcdirmid on Tuesday March 24, 2015 @08:14PM
(#9260483)
You don't, but so what? Even proofs done well by hand can have obscure mistakes; most of the time they don't invalidate the proof. Take a proof as strong evidence of correctness and it all works out from a epistemological point of view.