We're testing remote 3D rendering over WebRTC. We render the 3D world in the cloud. Video and controls go over a WebRTC-connection. I wrote a simple nodejs websocket server to handle the signaling.
It has been a quite wonderful experience despite the tech being somewhat experimental. Google Chrome offers excellent tools (chrome://webrtc-internals/ & chrome://webrtc-logs/).
First I wrote the client and a mock server to work in the browser while a colleague started implementing webrtc to work in the renderer with the PeerConnectionpp-lib. Debuggin was awesome, when you had a working mock renderer in the browser.
This is quite interesting. Does this mean all clients are seeing the same video rendering? What made you do it this way rather than render on the client and update positioning over WebRTC?
Regardless id love to read more about your progress if you have anything written down.
One use case for cloud rendering is low-end mobile devices that can't handle the rendering of complex 3d-scenes. Or you could create a live presentation using a 3D world, eg. about an architectural design.
Other people in the project are building tools on top of WebGL to do the rendering in the browser - mainly porting the [realExtend toolset](http://realxtend.org/) to Javascript.
Everything I do will be released as open source later on. I'll give you a shout when we got something to show.
XMLHttpRequest cannot load https://api.talky.io/socket.io/1/?t=1383140651742. Origin https://talky.io is not allowed by Access-Control-Allow-Origin. talky.io/:1
I am digging thru the logs now to find out - early signs indicate memory or resource leak in the api and it then silently crashed (well, silently from my side - i'm the Ops person for &yet)
I'm afraid I could not get the demo work when I click create room nothing happens. But it's nice to see more libraries popping up around this technology.
It has been a quite wonderful experience despite the tech being somewhat experimental. Google Chrome offers excellent tools (chrome://webrtc-internals/ & chrome://webrtc-logs/).
First I wrote the client and a mock server to work in the browser while a colleague started implementing webrtc to work in the renderer with the PeerConnectionpp-lib. Debuggin was awesome, when you had a working mock renderer in the browser.
All the bugs I encountered we're my own and I found all the answers I needed in the dev API doc: http://dev.w3.org/2011/webrtc/editor/webrtc.html