client

Explanations

1. User 1 triggers the method startAudioChat. He then grants access to his microphone and inform (through socket.io) User 2 that he is ready to initiate a peer-to-peer communication. His media stream is stored in a variable.
2. If User 2 accepts the invitation, he then grants access to his microphone and iniates a peer-to-peer communication by sending (through socket.io) the Offer (including his media stream) to User 1.
3. Step 3 happens automatically. User 1 receives the Offer and send back the Answer to User 1 (including his media stream that was stored in step 1)
4. Now User 1 and User 2 have all the necessary information to communicate with each other via peer-to-peer.

Note

Before starting the conversation via WebRTC, we want to check 2 things :

1. Does the user pressing the "audio chat" button has a microphone and allow the website to access it?
2. Does the receiver really want to engage into a audio conversation?

These two points are the reason why we only store the media stream and do not already iniate the peer-to-peer communication. Instead we wait for the receiver to accept the audio call. It then can seem counter-intuitive, but it is actually the receiver that iniates the peer-to-peer communication, since we know that user 1 want that communication and has already given access to his microphone. When user 1 receives the offer from user 2, we make use of the store media stream.

backend

https://github.com/mornir/node-chat-app

Resources used

WebSockets

Official Getting Started Guide Andrew Mead Udemy Course

WebRTC

Theory

[WebRTC and the mechanics of peer to peer networking](WebRTC and the mechanics of peer to peer networking)

https://www.html5rocks.com/fr/tutorials/webrtc/basics

Tutorials

Grafikart WebRTC

Simple-Peer

Server list

STUN: https://gist.github.com/zziuni/3741933 (free) TURN: http://numb.viagenie.ca/ (free)

Project setup

npm install

Compiles and hot-reloads for development

npm run serve

Compiles and minifies for production

npm run build