Programming with Types

I'm happy to announce the early access launch of my book, Programming with Types.

This is the culmination of several years of geeking out on type systems and software correctness. I've always liked to learn how to write better code, but if I were to point out exactly when I started down this particular rabbit hole, I'd say it was 2015.

I was switching teams at that point and decided to get up to speed on modern C++. I started by watching C++ conference videos and was mind-blown by the C++ Seasoning talk Sean Parent gave at Going Native a couple of years before. That gave me a completely different perspective on generic programming.

On a parallel thread, I was playing with Haskell and learning about the advanced features of its type system. Programming in a functional language makes it obvious how some of the features taken from granted in such languages get adopted by more mainstream languages as time goes by. Closures, sum types, and monads are slowly making their way to the mainstream.

Bibliography

In awe of the elegance of generics and trying to learn more about them, I picked up Stepanov's From Mathematics to Generic Programming and Elements of Programming. I realized I need a refresher on abstract algebra, so I took a detour and went through Pinter's A Book of Abstract Algebra.

Getting a sense of the mathematical underpinnings of generics made things much clearer for me. I wanted to get a similar understanding of the math underlying the Haskell type system, namely category theory. A great resource on category theory is Bartosz Milewski's Category Theory for Programmers.

Deeper still down the rabbit hole, I picked up Benjamin Pierce's famous Types and Programming Languages. The book covers many aspects of type systems, from basic types and function types, to subtyping, generics, and higher-kinded types. In fact, this is exactly the progression my own book follows. Types and Programming Languages is geared towards compiler writers. I wanted to write something that can benefit any developer.

From Theory to Practice

While learning more and more about type systems, I could tell the code I was writing at work became better. There is a direct link between the more theoretical realm of type system design and the day-to-day production software. This isn't a revolutionary discovery - all fancy type system features exist to solve some real-world problems.

I had new insights which I could use and did my best to share them. I started this blog and posted about various practical applications. I did hundreds of code reviews and applied what I learned there. And here is where I found my niche.

Not every practicing programmer has the time and patience to read highly-theoretical books, with mathematical proofs. On the other hand, I could tell that my time wasn't wasted reading such books - they made me a better software engineer. I figured there is room for a book which covers type systems and the safety features they provide, with practical applications anyone can use in their day jobs.

Programming with Types {#programming-with-types-1}

The book starts with basic types and some of their common pitfalls: numerical types tend to overflow or are subject to rounding errors, strings have several encodings, and manipulating them naively causes all sorts of issues. There are lesser known basic types, like the empty type and the unit type, which are not as popular for some reason, even though they have great applications: as return types for functions which never return, or don't return anything meaningful.

After basic types, the book covers composition. Why record types are generally better than tuples, what algebraic data types are about, and countless applications of sum types. Functions should return either a valid value or an error, never both. The variant design pattern, enabling double-dispatch, looks different today than it did a few years ago.

Function types are discussed at length, from lambdas and the functional staples map, filter, and reduce, to modern features of programming languages like yield and async/await. The book shows modern takes on the strategy and decorator design patterns, implemented more succinctly using function types.

Subtyping is another major topic, covering not only the elements of object oriented programming and how to use them effectively, but also variance, top types, and bottom types. For example, we can use a bottom type to produce a value out of nowhere. Mixins are controversial, as they are usually implemented as multiple inheritance, but I believe they are extremely useful when designed correctly.

The next major topic is generic programming. Generic data structures are responsible for shaping the data, while algorithms are responsible for processing data. Iterators are a bridge between data structures and algorithms, allowing us to mix-and-match them. As an example, we can find an item in a tree using the same code we use to find an item in a list.

Finally, the book covers higher-kinded types. These are higher-level abstractions, generic types with generic arguments, which underpin concepts like functors and monads. The joke goes that as soon as you understand monads, you lose the ability to explain them. I'm taking it as a challenge.

Check out my book here.