Reactive Programming with RxJS

Vinh Le
Vinh Le
Sep 24, 2019 · 7 min read

Part 1: Why RxJS?

During previous summer, one of my biggest achievement was to get the hang of RxJS and Reactive programming philosophy in general. Approaching these really calls for a different mindset. It could also be challenging to understand declarative programming while imperative programming is deeply in your daily toolset.

Therefore, I hope that throughout this RxJS series, we can go through ideas and real-world implementations together. Then who knows if this could be a game-changer for your project. So, let’s dive in the first part 🚀

RxJS in a nutshell

RxJS is a library for composing asynchronous and event-based programs by using observable sequences

Think of RxJS as Lodash for events.

Essentially, RxJS was born to solve problems that JavaScript developers encounter most notably in event-based and async actions.

However, for these two, Promise and Callback have been standard solutions for years. So why bothering to have a new paradigm? 🙄 And which use cases can we get the most out of RxJS?

In this blog, I will show you 4 primary reasons why RxJS is a perfect replacement for Promise and Callback. Now, get yourself some coffee ☕️ and let’s get into it 🚀

Handling different data sources at ease

getCakeButton.addEventListener('click', (e) => console.log('Give me cakes!'))

Another source of data comes from server/database level. We can use either Callback or Promise to handle this. Here is an example of handling API request by Promise:

const getCakes = async () => {
// Get cakes 🍰
const myCakes = await )
// Consume cakes 🍽

On the other hand, RxJS treats all sources of data as streams in an exact manner. These sources include reading a file, making an API request, clicking a button or moving the mouse. Let’s see how RxJS handle our previous button click and getCakes 🍰 API request:

// A stream of getCake button clicks
const source = fromEvent(getCakeButton, 'click')
tap(e => console.log('Be quick!'))
// A stream of fetchCakes promisesconst promise = fetch(cakeEndpoint, {method: 'GET'}.then(cakes => cakes)
const source = from(promise)
map(cakes => consumeCakes(cakes))

From here, we can see that RxJS does not necessarily reduce the amount of code dealing with these 2 particular problems. Nonetheless, it shows a consistent way of handling various types of event.

This actually calls for a shift in mindset. Thinking in stream is one key principle of RxJS. It treats all sources of data as streams. Besides, RxJS does not handle individual event but rather combining all of them to a stream and handling only values created by that stream.

If you look closely on RxJS codes above, it is apparent to notice source of the data. It could either be button clicks or API requests. All button clicks are combined into 1 single stream. RxJS does not care if user actually clicks it once, or twice. It treats all of those clicks similarly and care only about value emitted from each click. API request stream works in a similar pattern.

Minimise side effects

In a brief, when a function relies on data whose source is outside of its local scope, it is prone to side effects. Let’s inspect side effects in action by complicate our get cakes a little bit 🕵 ️‍🧐

Now imagine we are allowed to eat only 3 first cakes regardless of how many we asked for 😔. In this case, we need to have a state variable to keep track of the amount of cakes we already consumed:

side-effects.jslet consumedAmount = 0
const consumeCake = cakes => {
// check if we still can consume it 😱
if(consumedAmount === 3) {
// do something with cake ☕️
alert('This cake is great')
// update the consumed amount
consumedAmount = consumedAmount + 1

Our consumedAmount variable seems to be a good fit for above solution. However, it is barely optimal. We need to update this variable every time we consume our cakes. In other words, we mutated this state variable.

When more features arise and further complicate this tracking logics, our application is prone to bugs. Because our variable is exposed and mutated by other functions. In debugging, we also need to examine every single logics that mutated this variable. This could be a challenging and daunting task when we scale our application larger.

In other hand, here is how RxJS simplifies the puzzle:

map(cakes => consumeCakes(cakes)),

Dead simple isn’t it? By treating data source as a stream, RxJS skips the job of keeping track of each time we consume the cake. Rather than that, it only cares about how many cakes that the stream is allowed to produces. In fact, take operator helps this out and free out logics from side effects and potential issues 🚀

In truth, functional programming (FP) does not guarantee a program to be entirely free from side-effects. However, it aims to minimise them by locking variables within function scopes. Besides, FP explicitly shows where sides effects are performed. So when our application is buggy, we could expect to jump directly to those places to hunt those 🐛

If you want to learn more about functional programming, Functional-Light JavaScript is a good place to start.

Declarative coding style

Additionally, a rich set of operation functions (operators) are provided in RxJS. Each operator follows this notion of having no side-effects and being declarative. We therefore can have stream handler like:

const source = from(cakes)
tap(cake => console.log('This cake is fine: ',,
map(cake =>,
.subscribe(cakeName => alert('Here is your cake 🍰: ', cakeName))

In here, each operator such as filter, tap, map is declarative and free from side effects. They neither mutate the source nor depend on any out-of-scope variables.

This is a rock-foundation for even more complicated operations that we might need to add to existing program. This comes to the 3rd benefit of RxJS.

Powerful in managing complicated streams

Imagining an use case where we want to search for cakes instead of just clicking get cake button to get them. Our API supports for filtering by name, so we will do just that on client side. The flow will look like:

// Input handlers to take input value for searching // Only make an API request for searching after 0.5s since user stopped pressing a key // Retry 3 times if an API request fails // Cancel the search when press Esc or clear input // Handling errors by showing an error alerts

Now, please take a seat and spend some times brainstorming how to implement this feature. In our normal way, we will use:

  • Callback to handle each input onChange event
  • Promise to make API request
  • Having a variable to store the status of API request. And another one to store the amount of retrying when one request fails.
  • Callback to handle ESC keyup event
  • Another variable to decide if we makes API request. ESC keyup event or Input onChange event will mutate this variable to either true or false
  • A wrapper try-catch or Promise .then().catch() to handle error is challenging isn’t it? 🤔 What do you see from those? I see a tons of side effects and a tons of nested callbacks and another ton of helper functions which are very inclined to bring interesting bugs 🐛 to the program.

Let’s see how we could use RxJS to solve this at ease while ceasing side effects and potential bugs:

How RxJS handles a complicated data stream

✅ Non-existence of state variables

✅ Declarative functions in which each one takes care of 1 specific operation

✅ Side-effects free

Key takeaways

  • RxJS embraces functional programming principles including declarative and side-effects free functions. This helps our applications less prone for potential bugs
  • RxJS brings code readability to the next level

Thanks for reading this first part of RxJS series! Stay tuned for the next one 🤩

✍️ Written by

Vinh Le @vinhle95


Say Hello 👋 on





Shot code

where coding shots are served

Vinh Le

Written by

Vinh Le

A hustler, lifelong learner, tech lover & software developer @vinhle95; 👨🏻‍💻💪🔥🎾

Shot code

Shot code

where coding shots are served

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade