The "Internet", as originally envisaged, does not exist. It hasn't existed for some considerable time. The original image was for a hyper-redundant platform that could survive anything up to (and including) a nuclear war.
Today's Internet can't even survive a bunch of freeloaders stealing copper wire, or some drunk maintenance crew cutting through cables during road repair.
The "Internet", as Universities envisaged it, in it's next incarnation, is also dead. It was intended as a network by which academics could trade information quickly and easily, on a national or even global scale. UDP was safe for real-time, because the risk of lossage was small.
Today, I doubt there's a major pipe that isn't dangerously overloaded most of the time. Real-time scientific data transfers are impossible. Communication is so slow and broken, it became more profitable to chalk up the billions spent on the Internet as a lost cause, and build Internet 2 for the scientific work.
Once the Internet was sold into slavery - err, the private sector, it was intended to be used by home users for web browsing, home shopping, telecommuting, etc. Videoconferencing and Internet-to-Phone relays also became hot topics.
This incarnation also died. The private sector cut corners and cut costs. At the same time, web crawlers, "active agents", e-mail spam, and other bandwidth-soaking systems have caused network grey-outs across the Internet. These have been largely responsible for the fear that the Internet was about to collapse. The robots file and the specialization of the search engines prevented a total collapse, but it was close.
In many places, such as Euronet, videoconferencing was banned, along with high-bandwidth games. The capacity of the network was just too small, and the money anyone was willing to spend to fix the problem was even smaller. Banning anything that didn't fit proved much cheaper.
At the same time as all this was going on, ISPs were getting blasted. AOL was even taken to court, over what was practically a non-existant service. (AOL lost. Big time.) The bandwidth they had was far and away inadequate for the task. Their routers were puny. Their phone lines were negligable. System resources were a disaster. And this was true across the board.
A -tiny- handful of ISPs were able to cope, and these flourished, whilst the others floundered. AOL survived the famine times by sheer bulk. Many didn't, and became prey for AOL and others. The -practical- Internet age, though, was now a rotting corpse. Starved of resources, it had wasted away.
The current Internet is nothing like its forerunners. It doesn't cater to "getting things done". Rather, it is a large entertainment centre. Prawn is the big #1 seller. Other online stores aren't profitable, never have been, and probably never will be. The Prawn industry is. Market forces dictate that the big WILL crush the small. There's only so much capacity to go round.
Online banks? A dream that died, long ago. Even Microsoft's attempt, with Visa, was killed. Micropayments? A neat idea, ingenious, and utterly unworkable in a cut-throat climate. Secure on-line transactions? Fewer than 25% of all online shops use =any= level of security. Of those with SSL, fewer than 10% use 128-bit (which is the mimimum that is worth using, if you want to -keep- your card safe). NONE of them use client-side certificates (which you need to prove that the client actually owns the card they're using.)
Once the transaction is done, though, is your data safe? SETI@Home kept personal data in plain text. Can you be sure others do any better? And even if it's kept safely, will it be sold to others? And =what= will be sold, if it is? And how long will that information be kept? Many places -keep- credit card numbers, after transactions are complete, for "verification" or for "quick entry". A very inviting list, if the company, or someone in it, wants to make a fast buck.
The result is a network that is now under the total control of the media moguls and the prawn pushers. The Internet is, and will remain (for a while) simply an entertainment forum. It's value as a tool to get things done is limited to tiny segments, such as the various Open Source/Free Software projects. Just small islands, scattered in a LARGE ocean that is full of sharks.
Of course, this incarnation is unstable, too. The markets being reached have only a vapor value. Prawn is natural. So is a cat hurling a hairball. The market for the latter is unlikely to ever be earth-shattering. And once society matures past electronic pubescence, prawn will also begin to lose its appeal. Value comes from rarity and novelty. Once these go, the market goes.
What's next? I honestly don't know. I don't even know if the Internet will survive another collapse. The Rulers of the Backbone have become (in their own minds) a powerful Elite, with end-users disposable peons. ICANN is at war with the IETF. xDSL providers are at war with each other, and will likely all die. Cable companies are at war with the telecos.
The essence of an "Internet" is a federation of networks, working together, co-operatively, supporting each other in a peer-based system. As soon as you add ANY level of dissent into the system, it WILL collapse. Maybe not today, maybe not tomorrow, but the wound IS fatal. There is no cure, at that point.
Why should that be so? Simple. Because you have to depend on everyone you're connected to to play fair, if someone decides you're not in their idea of the future, you're in big trouble. ONE change to ONE router table, and your network ceases to exist. If that's too chancy, lower the priority of your network. The slow-down will make people leave. The network's then vulnerable to a hostile take-over.
Oh, yes, the Internet Protocols work very effectively to produce a QUALITY service, IF used to that end. They ALSO make deadly weapons, capable of killing any company, however rich or powerful, if wielded in the hands of someone sufficiently evil and/or ruthless.
Any time in the next decade, such a war is very likely to occur. We're already seeing skirmishes, with national filtering and "content filters" that don't just filter content. The war WILL break out, and it WILL spell the end of the Internet as we know it. Very little of what we know as the Internet will survive.
After the Great Internet War? Who knows? Whoever wins will own a rotting, misshapen, scarred infrastructure that might prove unsalvagable. And building another might be considered too expensive. Nobody today would consider building a full-scale Egyptian "needle", or a genuine "Stonehenge". The cost and danger would be prohibitive, and the benefits would never cover the cost.
In 2010, this, then, is my prediction. The Internet will no longer exist. It will not be replaced, with a network federation. Rather, small, specialised, low-cost networks will emerge, totally disconnected and totally isolated. Bridges -might- appear, as needed, but not on any grand scale. That will have gone forever.
The isolation will kill the web, completely. There's no need to have a presentation system over such small clusters. It will also kill many large, co-operative projects, such as Open Source and Free Software. These don't scale down well. Not all will go. Small, flexible, adaptable projects will survive and evolve. Compatiability will be lost, over time, but the code will go on.