Writing a real application in Haskell

Published on 5 February 2012.

24 September 2017: Fixed broken link to Mercurial.

In this post I explain why I choose to write the file organization application in Haskell.

Getting away from OOP

I’ve spent most of my time programming in object oriented languages, but I’ve also been interested in functional programming ever since I was introduced to Lisp at university.

It was in the Lisp course that I first learned about recursion (or at least came to understand it), and it also taught me a lot about programming in general.

The solutions that we wrote had a different structure from what I was used to. I remember them as powerful and elegant. Somehow that elegance never made it outside the university and the Lisp course. That makes me a little sad.

One big reason I find object oriented programs hard to understand is mutable state. Since that is less of a problem in functional programs, I keep thinking that functional programming must be better. Or at least yield less complicated programs.

The paper Why Functional Programming Matters also makes some compelling cases for functional programming.

So how come it’s not used mainstream?

Is it because it doesn’t work in practice or is it because it’s so fundamentally different that people are not willing to wrap their heads around it?

I decided it was time for me to find out by trying to write a real application in a functional language.

Why Haskell?

I wanted to write a real application in a functional language. But why Haskell?

I came across Haskell by accident. Because I was learning Mercurial, I was reading Mercurial: The Definitive Guide, written by Bryan O’Sullivan. I found out that he had also co-authored the book Real World Haskell. Since I liked his first book, I started reading the one about Haskell.

I found Haskell interesting for two reasons: you can obviously build real things in it (and I had a book to show me how), and it’s a purely functional language.

Building real things was important because I didn’t just want to play around in a REPL. Purely functional was important so that had to force myself to adopt the functional paradigm.

Another feature that makes Haskell different is lazy evaluation. It’s a feature that allows programs to be structured in a way that would otherwise be impractical. It’s also stated as one of the two advantages of functional programming in the paper mentioned above.

Conclusion

By writing the application in Haskell, I’m exploring a different way of structuring my programs. I’m expecting to learn a lot, and I will share what I learn on this blog.