Idris2

It is, and it still exists... it's also a brand new compiler, instead of one with nearly 30 years of development on it, and a brand new ecosystem, instead of having the massive amount of hours that Haskell has in it. Idris is really nice imo, but Haskell+Dependent Types will likely be usable sooner. There is an 'Idris 2' being developed, too (https://github.com/edwinb/Idris2), which is an entirely new compiler with a new type theory (although not entirely dissimilar), which just makes it even harder to use for real things (again, IMO).

I'm not an expert, but yes compile times can get crazy if you're not careful. In theory, the situation is no different from (ab)use of C++ templates, the Rust macro discussed here, and so on. In practice, whereas C and Rust will choke on a 100kloc function, Idris might choke on a 100-liner, depending on what you're doing. Edwin Brady is currently making steady progress on a new compiler--this time written in Idris itself--that's much faster in practice: https://github.com/edwinb/Idris2

So I use QuickCheck often and I'm a true believer, but the point this post is making is absolutely justified. Tying the test generation to the type is often requires some creative ways of writing `Arbitrary` instances (the specification of how test cases should be generated). However, I don't think what the author is suggesting in its place better. The advantage of tying the test case generation to the type is that it's compositional. Lots of sub Arbitrary instances come together to generate the test case and in the long term I find this to be more maintainable. Whereas the author's approach discourages compositionally. The real problem is that the types in vanilla Haskell are too loose, so there lots of inhabitants of a type that we are not interested in. The author mentions in passing, dependent-types might be a solution and that is exa

repository (published a few days before this talk at Curry-On 2019): https://github.com/edwinb/Idris2