As we continue to develop technologies such as Peercasting and PublishSubscribe, the real problem we are trying to solve is one that should have been solved years ago. The issue at hand is Multicasting.
Unlike broadcasting, wherein everyone gets hit up by the broadcast, multicasting only goes to those interested. The trick here is that the routers handle the multicast in the most efficient way so that there is no redundant bandwidth being wasted.
Problem is, most internet connections don’t seem to support Multicast. Sure, there is the Mbone, but I’ve yet to see an ISP support it. Thus, we have to do all this crazy stuff with stream relays and crap, just to do something that the Internet was supposed to do in the first place.
As it stands today, all we really have on the internet is Unicasting, connections from one person to one person. When you want to send something to two people, you have to waste bandwidth by sending the same data out twice.
Peercasting is only a temporary solution. What we are doing is esentially some P2P relaying of streams, which may help distribute the bandwidth waste, but that waste is still there. It should not be the end user’s responsability for Multicasting, it should be the routers’.
There is so much potential in Multicasting. With it, we could build entire new media distribution systems cheaply and efficiantly. As stated before, anyone with a webcam and a DSL connection could Multicast to millions.