26 November, 2019
What is WebRTC?
Web Real-Time Communications – WebRTC in short – is an HTML5 specification that allows you to communicate in real-time directly between browsers without any third-party plugins. WebRTC can be used for multiple tasks (even file sharing) but real-time peer-to-peer audio and video communication is obviously the primary feature and we will focus on those in this article.
What WebRTC does is to allow access to devices – you can use a microphone, a camera and share your screen with help from WebRTC and do all of that in real-time! So, in the simplest way
WebRTC enables for audio and video communication to work inside web pages.
WebRTC is a complex topic where many technologies are involved. However, establishing connections, communication and transmitting data are implemented through a set of JS APIs. The primary APIs include:
- RTCPeerConnection – creates and navigates peer-to-peer connections,
- RTCSessionDescription – describes one end of a connection (or a potential connection) and how it’s configured,
- navigator.getUserMedia – captures audio and video.
To make a remote connection between two or more devices you need a server. In this case, you need a server that handles real-time communication. You know that Node.js is built for real-time scalable applications. To develop two-way connection apps with free data exchange, you would probably use WebSockets that allows opening a communication session between a client and a server. Requests from the client are processed as a loop, more precisely – the event loop, which makes Node.js a good option because it takes a “non-blocking” approach to serve requests and thus, achieves low latency and high throughput along the way.
Do you have experience with microservice architecture? 🤔 Take part in the State of Microservices 2020 survey!
Demo Idea: what are we going to create here?
We are going to create a very simple application which allows us to stream audio and video to the connected device – a basic video chat app. We will use:
- express library to serve static files like our HTML file which stands for our UI,
- socket.io library to establish a connection between two devices with WebSockets,
- WebRTC to allow media devices (camera and microphone) to stream audio and video between connected devices.
Video Chat implementation
The first thing we’re gonna do is to serve an HTML file that will work as a UI for our application. Let’s initialize new node.js project by running:
npm init. After that we need to install a few dev dependencies by running:
npm i -D typescript ts-node nodemon @types/express @types/socket.io and production dependencies by running:
npm i express socket.io.
Now we can define scripts to run our project in
When we run
npm run dev command, then nodemon will be looking at any changes in src folder for every file which ends with the
.ts extension. Now we are going to create an src folder and inside this folder, we will create two typescript files:
Inside server.ts we will create server class and we will make it work with express and socket.io:
To run our server, we need to make a new instance of
Server class and invoke
listen method, we will make it inside
Now, when we run:
npm run dev, we should see:
And when we open the browser and enter on http://localhost:5000 we should notice our “Hello World” message:
Now we are going to create a new HTML file inside
In this file, we declared two video elements: one for remote video connection and another for local video. As you’ve probably noticed, we are also importing local script, so let’s create a new folder – called scripts and create
index.js file inside this directory. As for styles, you can download them from the GitHub repository.
Now, you need to serve index.html to the browser. First, you need to tell express, which static files you want to serve. In order to do it, we will implement a new method inside the Server class:
Don’t forget to invoke
configureApp method inside
Now, when you enter http://localhost:5000, you should see your index.html file in action:
The next thing you want to implement is the camera and video access, and stream it to the
local-video element. To do it, you need to open
public/scripts/index.js file and implement it with:
When you go back to the browser, you should notice a prompt that asks you to access your media devices, and after accepting this prompt, you should see your camera in action!
How to handle socket connections?
Now we will focus on handling socket connections – we need to connect our client with the server and for that, we will use socket.io. Inside
After page refresh, you should notice a message: “Socket connected” in our terminal.
Now we will go back to
server.ts and store connected sockets in memory, just to keep only unique connections. So, add a new private field in the
And on the socket connection check if socket already exists. If it doesn’t, push new socket to memory and emit data to connected users:
You also need to respond on socket disconnect, so inside socket connection, you need to add:
On the client-side (meaning
public/scripts/index.js), you need to implement proper behaviour on those messages:
Here is the
Please notice that we add a click listener to a user container element, which invokes
callUser function – for now, it can be an empty function. Now when you run two browser windows (one as a private window), you should notice two connected sockets in your web app:
After clicking the active user from the list, we want to invoke
callUser function. But before you implement it, you need to declare two classes from the
We will use them in
Here we create a local offer and send to the selected user. The server listens to an event called
call-user, intercepts the offer and forwards it to the selected user. Let’s implement it in server.ts:
Now on the client side, you need to react on
Then set a remote description on the offer you’ve got from the server and create an answer for this offer. On the server-side, you need to just pass proper data to the selected user. Inside
server.ts, let’s add another listener:
On the client’s side we need to handle
We use the helpful flag –
isAlreadyCalling – just to make sure we call only the user only once.
The last thing you need to do is to add local tracks – audio and video to your peer connection, Thanks to this, we will be able to share video and audio with connected users. To do this, in the
navigator.getMediaDevice callback we need to call the
addTrack function on the
And we need to add a proper handler for
As you can see, we’ve taken stream from the passed object and changed
srcObject in remote-video to use received stream. So now after you click on the active user, you should make a video and audio connection, just like below:
Now you know how to write a video chat app!