More Code Please
Since we began building OpenPath, there have been some notable improvements to WebRTC. Originally we were using webrtc.io as our WebRTC framework. However, PeerJS has gained momentum, and is simple, stable, and supported. It interoperates amongst browsers and browser versions, and is open source and free. This framework is helping us clean up our API calls and rationalize the backend. This will make the app faster, and also easier for us to build on and other coders to understand.
Recently, Mozilla also made the announcement that Persona, its simple sign-on service, would not be supported with developers in the future. One of the reasons for the move was that it didn't have a big enough footprint. We have to agree; further, people we tested with were confused by it. Sometimes it didn't work (often the user's email provider wasn't supported for instance). They didn't understand how it worked, who was asking for their personal info, and how their data would be stored (though Persona is designed to give users the freedom to decide what they share with services).
We think other authentication methods will present less friction; so we are migrating to passport.js. It is an authentication middleware for Node.js. It supports all common authentication strategies; we will look for input on which ones we implement as we go along. In case you are wondering, we have been working on a different branch of the code, which we have not deployed to the prod server. Of course, this is all in an effort to be able to pilot OpenPath and help people learn face to place!











