I was listening to Dana Blankenhorn on The Impacts of Persistent Distributed IT at Accelerating Change. He suggests that we do not have a platform for this decade and that wireless (wimax and wifi) could be it. He states the simple but actually quite obvious fact that we use wireless to access the internet. So everyone is cramming the pipe beyond the access point and not using the actual wireless in ways we are only just starting to see.
And honestly it all makes sense. Wireless has effective bandwidth from 3meg (802.11b) to 100meg (WiMax 802.16). Why not use the wireless point as a platform? Dana suggests Linux is the key for this and he's not wrong. For example, I saw this linksys wireless router ages ago which can be hacked to put linux on it. Yes thats great but wheres the use in just having linux on it? Yes I know you could install anything you like on it but besides a webserver I've seen little else. Till a while ago when the Xlink guys released Xlink Kai station for the WRT54G. Which means you can route traffic for playing multiple player games on the free xlink network with just a xbox and one of these. (yes i have talked about it before, but highlighting whats possible when you think about wireless as a platform).
Yes limited example but a interesting none the less. I would like to see zeroconf aka rendezvous better known as multicast dns services happening at the wireless level. I mean its ideal because the automatic discovery nature makes roaming around wireless points a joy. And before I go, lets not limit wireless to one point. With Mesh networking it should be simple to extend the range and the users to the service. Lets also not forget machine to machine services would benefit from mesh technologies.
