This paper both introduces a method for evaluation, which the paper's authors say is their primary contribution, and then use the method to identify some perfor

This paper both introduces a method for evaluation, which the paper's authors say is their primary contribution, and then use the method to identify some performance problems in Typed Racket. Fortunately, we've been able to make some of those performance issues a lot better just by fixing small things. Some of those fixes will be in the next Racket release, and some in the one after that. However, some of the performance overhead requires a more sophisticated compiler, just like higher order functions or objects. I'm working on this using jit compilers, and other people are looking at other approaches. I think this will be a major area of research in the near future. Full disclosure: I'm the author and maintainer of Typed Racket.