Rust is generally fast-by-default. My favourite example is the collections API, where you can do mapping/reducing/filtering operations only by inlining multiple

Rust is generally fast-by-default. My favourite example is the collections API, where you can do mapping/reducing/filtering operations only by inlining multiple operations and going through the collection only once. With Haskell these things are possible, but not efficient by default (the default Haskell library uses liked lists instead of vectors for example).