Because the problem solved by the Apple TV involves a branch in the process.
1) User is using their mobile device (happily) to locate content that may come from a wide variety of sources
2) User locates content that they'd like to share on a larger screen
3) User "hands off" (this is the branch) the playback of the content to the Apple TV
Many complex interactions might have occurred in steps 1 and 2. These interactions, typically, don't translate well to television screens. Based on adoption rates, no one wants to use a keyboard in their living room, and the content users want is already accessible on their mobile device. App developers welcome the opportunity to not develop for yet another device/platform.
So, you have two choices:
A) Let the mobile device do the heavy lifting and stream only "dumb" video to the Apple TV...
or
B) Devise some schema for handing off the entire session necessary for the Apple TV to connect (and authenticate, authorize, etc) to the media source.
Option B is full of pitfalls and difficult problems. Many are already solved, but they're just not there yet, IMO. I think we'll get there eventually though. As more and more services use the "Sign in using X" (where X is Facebook, Twitter, Github, etc) model, more and more users will become accustomed to federated login solutions. That is, ultimately, what would be required to hand off the direct connection to the media source.
Good points. I guess I just fall into the category of people who want my phone/tablet to act as a fancy remote control, more than I want to be able to transition from watching a specific thing between multiple places.
1) User is using their mobile device (happily) to locate content that may come from a wide variety of sources
2) User locates content that they'd like to share on a larger screen
3) User "hands off" (this is the branch) the playback of the content to the Apple TV
Many complex interactions might have occurred in steps 1 and 2. These interactions, typically, don't translate well to television screens. Based on adoption rates, no one wants to use a keyboard in their living room, and the content users want is already accessible on their mobile device. App developers welcome the opportunity to not develop for yet another device/platform.
So, you have two choices:
A) Let the mobile device do the heavy lifting and stream only "dumb" video to the Apple TV...
or
B) Devise some schema for handing off the entire session necessary for the Apple TV to connect (and authenticate, authorize, etc) to the media source.
Option B is full of pitfalls and difficult problems. Many are already solved, but they're just not there yet, IMO. I think we'll get there eventually though. As more and more services use the "Sign in using X" (where X is Facebook, Twitter, Github, etc) model, more and more users will become accustomed to federated login solutions. That is, ultimately, what would be required to hand off the direct connection to the media source.