While AI theories abound, where are the practical applications with solid business cases? They're no further than your current...
Run silent, run deep — VP8Chris Koehncke
There amongst the vendor hyperbole at the Atlanta WebRTC World stood Google. They don’t say much, they don’t speak loudly but we should listen to what they’re saying. It’s a signal, a sign, so don’t drive by it too fast.
As the arguments about codecs rage, Google, nearly in passing said, “we are working with mobile chipset manufacturers to incorporate the WebRTC codecs (meaning VP8 and Opus).” Immediately all sorts of folks start running around screaming “H.264 is the standard, what are they trying to do?” Well AM radio used to be the standard. And your point?
Embedding codecs in chipsets isn’t a trivial pursuit. It’s a long pull process from design to fab and along the way there has to be a commitment for large scale volume.
While Google is mostly quiet, there is no point in making noises about things to come. However, we all recognize for WebRTC to truly be a big deal, it has to work on a lot of mobile devices and since video is a big drive and factoring in the largest video content provider is in fact Google. You don’t have to be terribly smart to connect these dots.
For now though there is the discussion of transcoding. Vendors are already busy erecting toll bridges along this route with a troll ready to emerge and collect. Transcoding is always horrid. A reduction in quality with an increase in expense and complexity. It’s a necessary evil but as you look to your own WebRTC applications, try not to make transcoding part of your Interstate highway.