Topic: Possible futures
So, as we move to put the back-end systems into a (ideally temporary) low-burn state, we should start to discuss what the ecosystem should be like in the future.
One solution is to simply bring back up a version of the original system pretty much "as is", but streamlined to reduce the server load.
Another solution is to build a system that is *spiritually* like the old system, performing many of the same functions - the ability to choose, configure and deliver widgets to the devices, as well as updates to the firmware and Control Panels, and allow authors to upload widget - but in a different way.
One solution is to replace the widget engine, which is currently Flash-based, to something more contemporary and more Open - options include HTL5 with Webkit, Qt, maybe some reduced version of Android/dalvik, or perhaps something homegrown.
One solution might be to make the entire network peer-to-peer, with a flexible "root node" system to allow for transience.
There are strong (and somewhat evangelical) advocates that want to see the devices completely divorced from an on-line service altogether. There are such solutions out there in various forms and with various levels of required technical knowhow, but it might be nice someday to craft one that is more tailored for this type of usage rather than a hack to trick a system designed for online use.
Each of these comes with costs (money, time, skills) that must be borne by *someone*, and many other tradeoffs.
There are probably others I've missed....ideas?