Show HN: A formal proof of deMorgan's law in lean
1–6 of 6 posts
Re: Show HN: A formal proof of deMorgan's law in lean
#2Formal proofs are becoming easier. This proof was achieved in a couple days of intermittent effort, starting with no knowledge of formal proofs.
Re: Show HN: A formal proof of deMorgan's law in lean
#3Formal proofs are becoming easier. This proof was achieved in a couple days of intermittent effort, starting with no knowledge of formal proofs.
Wouldn't easier usually mean shorter proofs and better proof automation? What is Lean offering over other proof assistants?
Re: Show HN: A formal proof of deMorgan's law in lean
#4Formal proofs are becoming easier. This proof was achieved in a couple days of intermittent effort, starting with no knowledge of formal proofs.
Wouldn't easier usually mean shorter proofs and better proof automation? What is Lean offering over other proof assistants?
By easier I mean there's a good tutorial and a low barrier of entry (you can learn to type proofs in the browser). When a couple years ago I decided to teach myself coq, I very quickly gave up, because there just wasn't any good entry-level resource available online.
Re: Show HN: A formal proof of deMorgan's law in lean
#5Earlier quoted context omitted.
Wouldn't easier usually mean shorter proofs and better proof automation? What is Lean offering over other proof assistants?
By easier I mean there's a good tutorial and a low barrier of entry (you can learn to type proofs in the browser). When a couple years ago I decided to teach myself coq, I very quickly gave up, because there just wasn't any good entry-level resource available online.
I strongly agree there's a lack of text along the lines of "theorem proving for regular programmers". The majority I ever read assumed a strong background in maths, logic and functional programming and were really off-putting.
Re: Show HN: A formal proof of deMorgan's law in lean
#6Earlier quoted context omitted.
By easier I mean there's a good tutorial and a low barrier of entry (you can learn to type proofs in the browser). When a couple years ago I decided to teach myself coq, I very quickly gave up, because there just wasn't any good entry-level resource available online.
I strongly agree there's a lack of text along the lines of "theorem proving for regular programmers". The majority I ever read assumed a strong background in maths, logic and functional programming and were really off-putting.
Well, have a look at the two tutorials to which I link in readme.md of the github project. The first is very suitable for programmers with minimal prior knowledge of maths, the second is aimed at the opposite audience (mathematicians without much CS knowlede) (and I think is actually even simpler).