3 ms·

I think this is exactly what's happening, but through the back door: rather than everyone switching to Haskell, you see features pioneered by languages like Has

I think this is exactly what's happening, but through the back door: rather than everyone switching to Haskell, you see features pioneered by languages like Haskell being implemented in more mainstream languages. Also, do not forget Haskell's roots were to create a programming language environment where people studying compilers could easily experiment with new features. Hence Haskell's large number of language extensions.

theseoafs 11y ago
Right. What I'm saying is: how long do these dependently typed language features have to "incubate" before they make it into Java? Because implementations of dependently typed languages have existed for as long as Haskell has, and we've known that it's been possible to capture "proofs" in a type system for almost a century. So how much longer do you expect it will take? Or, maybe, there just isn't a real industry demand for dependent types. Hell, many engineers still have trouble convincing their employers to write robust tests, let alone full proofs of correctness.