webrtc-for-the-curious VS mediasoup

Here I will briefly go over the basics of how Web RTC works; for those who are interested in going a little deeper, I’ll leave the link here. In order for two peers to be able to provide themselves with RTCPeerConnection, the SDP (Session Description Protocol) protocol is used. The protocol has a key-value structure and is essentially a description of a single peer (the name speaks for itself).

I have worked four jobs related to https://github.com/pion/webrtc and one for https://webrtcforthecurious.com

Two companies used Pion. The other two were just using the protocol (WebRTC)

For the WebRTC jargon check out https://webrtcforthecurious.com/

If that still doesn’t cover enough I would love to hear! Always trying to make it better.

It is pretty easy to get a one way trip time for packets that is sub-second! You see it with conferencing and other real-time communication things.

If you are curious on the 'how' of WebRTC I wrote a Free/Open Source book that goes into the details https://webrtcforthecurious.com/. Happy to answer any particular questions you have.

Read this thoroughly: https://webrtcforthecurious.com

Given the financial restraint, i was avoiding paid API's like twilio. I started looking at mediasoup https://github.com/versatica/mediasoup, but while implementing the SFU server, its seems a lot more involved. For ex, TURN and STUN, peers negotiating different video codecs, adaptively changing the quality of video etc. Is it usually this difficult to implement a video conferencing apps ?

This release ships lots of new features to the already comprehensive set of them. Currently, we offer several working tutorials on how to set up STUNner with widely used WebRTC media servers and other applications that use WebRTC in Kubernetes, such as: - LiveKit - Jitsi - mediasoup - n.eko - Kurento

Architecture WebRTC SFU (server with Selective Forwarding Unit). Can handle unlimited rooms without limits of time, each having 8+ users, potentially many as it is scalable. Routing is a multiparty topology, where each participant sends its media to the MiroTalk media server mediasoup and receives all other’s media from it. This version is Ideally suited for large group video conferences.

I started the MiroTalk P2P & MiroTalk SFU projects during the pandemic period (about 1+ year ago), not knowing anything about the WebRTC. Making often the video conferences with my colleagues and not wanting to depend on Zoom, Teams, Google Meet... I decided to do some research about how it works and from there MiroTalk was born :) I Giving to everyone the chance to have its own instance of MiroTalk, which can be customized as you like and run in any cloud, vps, server. If you're just starting out, I suggest you take a look at the MiroTalk C2C (New) code, which can be a good starting point to understand how the architecture WebRTC Mesh (P2P) works. Later you can also study how the WebRTC SFU (Selective Forwarding Units - I recommend mediasoup which I personally love) or MCU (Multipoint Control Unit) architecture works. I wish you all the best!

The "Mediasoup" project provides a high level JavaScript/TypeScript interface to the WebRTC APIs. The core logic of this project is implemented in C++/Rust. Consider taking a look at the project if you want an easy-to-use library instead of the low-level libWebRTC APIs. A notable project to mention is the Pion/webrtc project which has a Golang implementation of the WebRTC API. Of course, we should mention the rust port WebRTC.rs. Let’s keep all the rustaceans happy too!

Indeed, maddening, especially as the wonderful https://mediasoup.org/ is developed here. Europe will never have great tech companies when the answer seems to be throwing €€€ away instead of investing locally

Hello thedominux, Thanks for your interest in MiroTalk ;) MiroTalk SFU code is: - JAVA-SCRIPT: 85.2% - HTML: 10.0% - CSS: 4.5% And has built in mediasoup server, more details about: https://mediasoup.org/

Mediasoup