[leoc]

I haven't been following closely, but my impression is that the dependently-typed languages/theorem-provers have stolen some of Haskell's "most purely functional, most strongly typed" hype-thunder even though they're even further from being a mainstream implementation language for most projects.