> The functional people have been able to hammer Rust into a functional language, sort of. Now they want a Haskell-like "do" construct so they can get imperativ

> The functional people have been able to hammer Rust into a functional language, sort of. Now they want a Haskell-like "do" construct so they can get imperative sequentiality back. Is that right? Not really, in that even before the blog gets to do notation, it depends on features that don't exist yet: generic associated types, which are RFC-accepted but not yet implemented, and associated traits, which haven't even been proposed. For now the whole thing is just a sketch. However, since associated traits aren't critical to the design (and would be a good feature to add anyway), at least some version of it will presumably be implementable in the future. Overall, Rust has always been inspired by functional languages to some extent, e.g. Rust traits are based on Haskell typeclasses, and Rust has pattern matching and full type inference. But I don't see `Monad` ever becoming idiomatic. I also don't see `do` notation being accepted, because its use cases largely overlap with other features, namely async/await and generators. On the other hand, it would require only minor changes to generators to allow implementing `do` notation on top of them, something I'm highly interested in mainly just because I want to prove it's possible. ;p