This page is maintained by Stephane Roux
TL;DR: Building Wonder is a complex tech challenge. Here's why!
- Hypergrowth with half a million users
- Live data transfer between thousands of users
- Highest performance requirements
- Rendering 2D spaces in WebGL
- Building our own WebRTC backend
Here are just some of the reasons why building Wonder is so complex:
Hypergrowth with half a million users
We're doubling almost each month and we now have around half a million monthly users. Scaling the system is a huge architectural challenge.
Live data transfer between thousands of users
We're transmitting live data (video, audio, text, images) between thousands of users simultaneously who meet and walk on Wonder, including real-time video processing.
Highest performance requirements
How much tolerance do you have for lag, connectivity-issues, downtime or poor audio quality? Probably zero. Wonder needs to be always up-and-running and milliseconds make a huge difference to the product experience.
Rendering 2D spaces in WebGL
While users move around Wonder, we're rendering the space in WebGL.
Building our own WebRTC backend
We're pushing the boundaries on what you can do in your browser. We're building our own backend in WebRTC and are working with some world-leading experts to satisfy the highest performance standards in a relatively new technology.