First, a reminder: the OS, if correct, must do all that it does easily for all users. Easily means easy for the user - everything's easy for the OS.
Basically, users want to stream video pretty much continuously at all times from assorted devices. Here streaming video means their device streams video to their memory. The default mode is streaming. Properly configured devices should stream for many hours without pause or attention.
All devices should of course be wireless. Memory is not necessarily entirely portable. We can call the non-portable component of a user's memory root memory, which is for the purpose of storing all of a user's continuous stream forever. Video streams to root memory wirelessly when the cameras are in range for that function. When cameras are out of range for that function they stream to portable memory components and those stream on to root memory when they are in range for the purpose.
Once video has been captured it needs to be reviewed, edited, and published. These capabilities must be transparently available to all users.
This component of the OS can presumably be assembled using existing technology, by individual users, so creating the component is not necessarily a device or streaming/editing dev project. It could be an ed dev project.
