The main concepts of functional reactive programming

Note

So today I watched a talk on functional reactive programming (FRP), which did a good job at explaining the basics of the pattern. FRP provides abstractions for values that change over time in a program, while keeping track of the time explicitly, as well as the stream of events that lead to that value changing. FRP is a good example of how the expressiveness and composability of functional programs can model real world interactions in an elegant way.

The main abstractions in FRP are behaviours and events.

1. Behaviours

   A behaviour can be thought of as state that changes over time. Semantically a Behaviour a is a function of type Time -> a, which can return its value at a particular point in time.

2. Events

   Events are a stream of state changes that affect a behaviour. They can be semantically thought of as pairs of a time and a value, a stream of events has type [(Time, a)].

So how do we use them?

We use them as any abstraction in FP - through combinators, building up big programs from smaller pieces.

Through the power of combinators and these functional abstractions we can separate the concrete implementation and execution from the description of the program, which lets us think about problems at a higher and more declarative level.

More to come later!