A formalization in Coq of the Haskell pipes library | Hacker News Reader

A formalization in Coq of the Haskell pipes library

github.com

73 pointslelf6 comments

I liked the little summary at the end:

> So in all it took 1.5 years to learn Coq well enough and to find the right abstraction, and 2 weeks to do the actual work.

That matches what other people have told me about Coq: the learning curve can be brutal, but you can be surprisingly productive once you get over it. (To be fair, this sort of problem also feels like a really good fit for Coq—proving theorems in more developed sub-field of mathematics is much harder.)

A pretty cool project all around!

I've had only a little exposure to Coq, but for the life of me I could not understand how to use it. I mean on the practical level, not the conceptual level.

I know people talk about it as a theorem-prover, but how do I actually _use_ that? This might be due to my general lack of exposure to theorem provers in general, but I can't wrap my head around how I can use this to make my programs better in a practical environment (despite my total buy-in to the concept)

Check out http://adam.chlipala.net/cpdt/

Also (and IMO this one is a little more well-structured) the Software Foundations course[0]. Especially when supplemented by Pierce's excellent lectures, available on youtube.

[0]: http://www.cis.upenn.edu/~bcpierce/sf/current/index.html

This is why I love the Haskell community! Awesome work!

Just glancing over this (super cool, John!), but what's going on here:

    Notation "f <~ g" := (f ~> g) (at level 70).

Seems likely to cause confusion... what's the reasoning/explanation behind this? I might imagine a law such as Notation "f <~ g" := (g ~> f), but...?

It's been too long since I wrote some Coq! Gotta get back into it :)

Reply on news.ycombinator.com