It just dawned on me recently--the endpoint as we know it is unimportant. And we have Google and WebRTC to thank for that.
If you deal with VoIP, then you are comfortable with either the term "Endpoint" or "User Agent." A few years ago, it usually referred to some physical device dedicated to calling--voice or video.
Later on, when software started eating up our world, endpoints were applications we installed on PCs to run calls.
And now with WebRTC, what does an endpoint really mean?
When Google's Justin Uberti stands at the Santa Clara WebRTC conference and announces over 1.2 billion enabled WebRTC browsers, he doesn't really mean endpoints. He means potential endpoints.
WebRTC also means that endpoints don't matter anymore. The endpoint is now the browser, which brings us to the end of an era.
It isn't about semantics--it is about focus. When the browser enables such interactions inherently, the use cases can change drastically. Best practices and beliefs around identity and federation come into question. The barrier of entry to development lessens, as does the friction complicating the adoption and use of a service.
We no longer talk about endpoints, but rather, services. And these aren't UC or pure VoIP services, but rather, higher-level services with voice and video used internally as a capability--as a feature.
The ramification here is twofold:
1. Business models relying on endpoints need to be reevaluated. Most of them probably will be strained in coming years. Pricing by subscription, or better yet by the conversation's context, makes more sense now.
2. With a footprint of over 1 billion potential endpoints, the game is about virality and real use of a system, and less about downloads and users.