functional stuff is making a comeback. haskell can be daunting in its pure-ness sometimes, requiring monads for seemingly anything useful. but it's a wonderful language/lifestyle choice. this online book broaches the subject pretty well: http://learnyouahaskell.com/chapters
If you want to go down the Functional rabbit hole more, "Purely Functional Data Structures" by Chris Okasaki would be a good next step. There's a book and a paper - from what I hear, the paper is just Okasaki's original work, whereas the book is a more complete coverage of the field.
To steal words from a comment earlier today, Haskell has a very steep but very rewarding learning curve.
It'll reinforce some concepts from math, and lead you toward writing code in a manner that's considered good practice in other languages - such that it's generally maintainable and testable. Similarly, if you're familiar with Design Patterns, you'll see that a lot of patterns are attempts to implement functional concepts in OO languages. That said, as a fan of Haskell I'm definitely biased, and a lot of that can be said for other functional languages as well.
"Real World Haskell" is also very good, but probably not the best place to start if you haven't written in a functional language before.
Plus learning a language from another paradigm, e.g. functional or logics programming, can also be insightful to your 'safe areas'[1]:
- After learning functional programming I realized how much of my programs do just mapping or folding over data. You start to spot the cases where this is not immediately obvious after some FP practice.
- After learning logics programming, I realized how much of my programs do a poor man's version of backtracking.
Maybe one of those languages is so much fun that you'll eventually change your 'lifestyle'. E.g. I use Haskell in many cases where I'd previously use Python or C.
[1] Of course, it doesn't always come natural. C++ STL does provide maps and folds, but since pre-C++0x doesn't have lambdas, using them requires more effort.
