Keith Watanabe * NET 2.0

I was chatting over email with a friend and hit an idea that I don't think anyone has really been pushing.  The idea comes from the basis that I've been saying the network has become the clear bottleneck in computing.  We've managed to make most of the other parts of infrastructure and software more redundant through cloud computing techniques.  However, the real killer that people are talking about is the upcoming limits to capacity in terms of the infrastructure.

Well, I don't see cloud computing limiting us since major companies like Amazon, Google, Yahoo, Microsoft and Sun are all proposing and providing means for allowing us to quickly shift our services onto their platforms.  However, I do see limits where bandwidth restrictions are imposed either artificial or just incompetence.

In a country like the US, logistics have played a key political point in deciding how infrastructure is laid out and has forced many people to be enslaved to a far slower internet compared to countries like Japan or Korea.  Personally, I feel that this limit is simply an excuse by the telephone monopolies and cable oligarchies to keep people bound to their crappy services.  But that's the supposed truth that happens whenever you try getting ADSL in the states.

So it's clear that physical lines are the bottleneck of more internet services and rich media from springing up on the net.  Imagine, for instance, if we didn't have to optimize for the web our pages if we had something more efficient in terms of allowing us to download.

Well, here's my solution: physical, wireless P2P hardware networks.

I'm not a network guy, nor will I ever care to be one.  But the thing is that common sense seems to dictate that what we did with P2P software, we can do with hardware.  Imagine this scenario.  You having small, wireless nodes installed in your home, on top of your roof that bounces signals to your neighbor and their neighbors all the way across town.  If a lag is detected because perhaps the signal died or it's down temporarily, another bounce occurs to find a new node.  I think this will work really well in large cities and probably be cheaper in terms of installations than digging lines under homes or city infrastructure.  I've seen certain ideas like this but the use seemed more like for businesses or security than home users.

In some way, this idea is kinda similar to ham radio, except that you're essentially mass sharing your signal with everyone.  Of course, you're going to face three major issues in this infrastructure format:

• The hops between cities and dealing with that.  Naturally, you'll need something to interconnect cities together.  Satellite or major ISPs would have to do some work in this regard.  Either that or perhaps some mass dishes that would provide major bounces between silent points (where silent points are like the edge of LA to Central California in Santa Barbara).
• The adaptation of technology to make it worthwhile.  For a service like this to work well, you'll need a lot of people subscribing to it or implementing it.  So perhaps universities and geek towns will be early adopters, but it'll be hard to get those AOL subscribers out in Arkansas to switch over ;)
• FCC and other potential regulatory bodies.  The FCC likes to fuck  things up like this.  But at this stage, who are they really benefiting in the long term by denying a service like this only to cater to the communication monopolies?

The main thing is that we need to move towards 1 TB/sec bandwidth.  That should be the goal in the next 1-2 years.  Cable modems and ADSL speeds in America aren't going to cut the next generation of rich media and internet applications to really push the web into overdrive.  For this to happen, we need something more efficient than the current crappy routers, hubs and ethernet that we're bound with.