Haskell High Performance Programming

Summary

We began this chapter by looking at what Haskell projects usually consist of and how Cabal and stack are used to manage project complexity and dependencies. We glanced at the basic usage of main test libraries and frameworks for Haskell and how they can be integrated into a cabalized project. We learned how to handle errors and exceptions. Even more importantly, we learned how to not do errors; why prefer throwIO over error (or throw)? Why are asynchronous errors so vicious in lazy semantics?

In the latter part of this chapter, we explored some Haskell trivia and techniques specific to GHC: lazy patterns, coding with GHC primitives (the magic hash), inlining, writing rewrite rules, using phantom types, fundeps, type families, the monomorphism ...