Ted Lemon did a very good job of describing the current state of naming models in Home Networking (homenet). This is a quite complex problem because it has at least two goals; and they may not entirely converge.
Goal one is to implement a model of ‘local’ information that we can call home.arpa — the things that are inside your own network. You don’t want to have to go outside the local net to find your fridge, or your TV or your burglar alarm on-off switch — you want these things to be local. Really local.
But there is this second goal: you need to find things out in the real world. You need to get to the real world of DNS names.
So, two goals — that’s two problems, right? Well no. It’s actually three (or more).
Because lying underneath is a third thing: you may have more than one provider in this modern world of homenet. You can consider it like being on a phone with both Wi-Fi and cellular, but it’s more complex than that because usually, when you have Wi-Fi and cellular, the choice of “which do I want to use” is well understood. It’s Wi-Fi. It’s usually faster and cheaper.
But, this actually masks what’s going on. What if your cellular provider has content for free, off bill, off quota, and your Wi-Fi is cheap, fast but also data count limited? What if using your Wi-Fi to watch Netflix means you can’t rely on it later on?
The problem about “which” service to use becomes a naming problem. So now we have
- Is it inside?
- Is it outside on provider A?
- Is it outside on provider B?
And the question — which provider do I really want to use, for this content? — well … it depends. It’s complicated.
Ted is trying to help people in homenet realize just how complicated this problem is. It takes name-to-address lookup, which is what the DNS was designed for, and it moves it into a much more complex space: names, to addresses, ranked in order of preference, depending on what you want.
We’re now talking about something new: brokerage. Brokering the “right” choice, depending on the situation you find yourself in.
Some comments in the room went to this being a bit of a complex problem, and emerging solution: it has many moving parts. I think thats true, but I think the moving parts are maybe as small as they can be, which is something Andrew Sullivan, Director of Dyn Inc’s Labs, said to the microphone much more cogently: the problem is complex enough to demand quite a lot of actors to solve.
The views expressed by the authors of this blog are their own and do not necessarily reflect the views of APNIC. Please note a Code of Conduct applies to this blog.