With our fourth WebRTC Conference & Expo now a week behind us, we wanted to take some time to reflect on the event. There was never a dull moment for Team TokBox in Atlanta. We gave a keynote address, participated in several panel discussions, gave a live demo of our video driver and manned our bustling booth complete with a Bridgestone Golf B-FIT kiosk.
While WebRTC has been innovating at an impressively rapid rate, the users of the web and mobile apps have been delighted with lots of new experiences. We’ve started connecting to people across different timezones, countries, and even continents in real time, on just about every sort of device. But when we ask developers, the people who dream up the next wave of crazy ideas, what they need in order to keep delighting their users we hear a few things over and over.
One of the most requested features of the platform that developers are patiently waiting for is WebRTC broadcasts at scale. The technical challenge is about getting the right stream (with the right bitrate, and the right encoding) out to all the different types of people who are watching, with their vastly different networks and bandwidth capabilities.
Hello! Ed from the BD team @ TokBox here. We’re always thinking of great ways to showcase cool partners, so we came up with an idea for a series called PartnerTok. This whole series will be done via our open source chat tool OpenTokRTC and recorded with our archiving stack! For our inaugural episode we are featuring our friends at Cambly. They’re a language marketplace for people who want to learn English or Spanish. We talked to them about where the idea came from, how it got started, their business model as well as their experience launching the app.
Cambly is also one of the partners testing out our new archiving stack – you can hear from them firsthand in the video about how easy it is to implement. In fact, we used the OpenTok API for WebRTC to power the live interview, and our new Archiving & Playback beta to record it.
- Archiving—You can record audio-video streams in a session and download the recording as an MP4 file (with H.264 video and AAC audio).
- Dynamic frame rate control—This feature lets you reduce bandwidth usage of a Subscriber’s video stream. This reduces CPU usage and the network bandwidth consumed, and it lets you subscribe to more streams simultaneously.
These are just a couple of the new features to be included in version 2.2.
In the last year we’ve witnessed VP8 proponents and H.264 proponents debate which codec should become “official” for WebRTC. The main points of contention? Licensing fees associated with H.264 make it unaffordable for a non-profits like Mozilla to support. In addition, VP8 isn’t compatible with existing and legacy video conferencing platforms which are typically built to support H.264.
We saw Google draw a line in the sand early on by announcing the “perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable” licensing of VP8. In addition, they recently moved their flagship video conferencing product, Google Hangouts, on to VP8.
Yesterday, Cisco unexpectedly announced that they will release an open-source version of the H.264 codec. The open-source version will include a free downloadable binary module that can be integrated into any application. All without the cost of licensing the codec . This is a strategic precursor to the IETF #88 next week where a vote will take place about the MTI (mandatory to implement) video codec for WebRTC, with the dominant front-runners being VP8 and H264.