No, he seems to get tripped up by the open world assumption, which is a limitation of Haskell's type system. It allows separate compilation of different modules

No, he seems to get tripped up by the open world assumption, which is a limitation of Haskell's type system. It allows separate compilation of different modules, for example. But nobody said that Haskell's type system is without limitations, right?