We've already talked about _The Reasoned Schemer_ in another thread. :) I may give it a try again eventually. Clocksin's _Clause and Effect_ is a similar book, in Prolog.
Monads aren't that hard, they're just one step too general - at first, it's not clear what their examples have in common. It helped when I realized I'd already written "monadic" code in OCaml and Scheme, and with pipelines in shell scripts. (Getting away from the avalanche of category theory terminology mixed with nonsense about burritos and space suits also helped. The epiphany itself matters, not the stepping stone!) Also, starting with a simpler monad (e.g. Maybe) makes the plumbing clearer.
For interpreters / compilers, I prefer the treatment in EoPL to SICP's. There's an awesome "converting an interpreter to a continuation-passing-style interpreter to a CPS-based compiler" section in (only) the first edition of EoPL.
What really impresses me about CTM is how much deep material it ties together, from all over the place, yet the book is still extremely easy to follow. Somebody just starting in CS could benefit tremendously from reading its take on concepts when their other textbooks stump them. It's one of the most readable textbooks I've seen. (PAIP is another.)
Monads aren't that hard, they're just one step too general - at first, it's not clear what their examples have in common. It helped when I realized I'd already written "monadic" code in OCaml and Scheme, and with pipelines in shell scripts. (Getting away from the avalanche of category theory terminology mixed with nonsense about burritos and space suits also helped. The epiphany itself matters, not the stepping stone!) Also, starting with a simpler monad (e.g. Maybe) makes the plumbing clearer.
For interpreters / compilers, I prefer the treatment in EoPL to SICP's. There's an awesome "converting an interpreter to a continuation-passing-style interpreter to a CPS-based compiler" section in (only) the first edition of EoPL.
What really impresses me about CTM is how much deep material it ties together, from all over the place, yet the book is still extremely easy to follow. Somebody just starting in CS could benefit tremendously from reading its take on concepts when their other textbooks stump them. It's one of the most readable textbooks I've seen. (PAIP is another.)