The Internet: Let’s call the whole thing off

| Apr. 15, 2007

According to this article on Yahoo! News, Researchers are exploring scrapping the internet.

Because the one we have now isn’t working out? I wasn’t really sure what brought on a need for a new internet, but some researchers think that we need a new internet because the one we have now barely works. But it isn’t too long before the article gets to the more cynical reason:

The first time around, researchers were able to toil away in their labs quietly. Industry is playing a bigger role this time, and law enforcement is bound to make its needs for wiretapping known.

This is exactly why a new internet is a bad idea. Just like including DRM in digital music at the behest of the RIAA, and the MPAA enforcing region codes on DVDs, the telecoms and wiretap-happy FBI are pushing for new features without any regard as to whether the end user actually wants them. And why should the average internet user trust the government and telecoms to have their best interests in mind? Anyone remember Net Neutrality?

As for the proposed improvements coming with this new internet, the details are scant and hardly the killer app that’s worth junking the current system over. And the cost of replacing legacy software and hardware to support this new system could add up to billions of dollars. No doubt ears perk up at AT&T, Pac Bell and Cisco Systems over the prospect of getting paid to rebuild the whole internet. So what do our billions buy us?

The new internet’s cheerleaders tout better security. Unfortunately, the threat of DDoS attacks, zombie PCs, spammers, hackers and getting your lil’ box pwn3rz0red don’t come from “The Internet”. They come from the security holes in the operating system from Redmond. We fix that and suddenly a lot of keyloggers, botnets and pop-up-ad-spewing spyware programs are rendered useless.

The new internet aims to “be redesigned to be skeptical of all users and data packets from the start. Data wouldn’t be passed along unless the packets are authenticated.” But this is happening already. The reason spammers and… well, pretty much just spammers, get away with sending out false email is because they usurp the rules of the current email system. Now they fake the ‘from’ address. In the radically new system, they’ll fake the from address and the authentication.

The article further assumes “faster computers today should be able to handle the additional [authentication] processing required within the network.” Hey, I thought this new system was designed to make things faster. Why not bog down the system with a few more security features while you’re at it? As it was for Alice in Wonderland (with apologies to everyone who’s sick of Wonderland/internet analogies), it’ll take all the running you can do just to stay in one place. And if you want to actually get anywhere….

Increased mobility is another proposed feature. The problem being that if you move your laptop from one location to another, you have to acquire an new address for it, which — unless I’m wrong — is the way addresses have always worked. I don’t know how mobile these people are planning laptops to be, but when moving a laptop from one wireless hotspot to another, it’s usually turned off anyway. So what’s the problem. If it isn’t, there is already zero-configuration networking, meaning your computer can jump from one network to another with a mouse click. The only data disruption you’re likely to face is by actually being out of range of the network. And unless the new internet can overcome the laws of time and space, that’s still going to be a problem.

Unfortunately, nowhere among the hand waving over security, mobility and… okay, the two reasons for scrapping a 40+ year old network, does the article mention improvements in speed. The internet was designed to send ‘packets’ of data, very small ones in fact, which made sense when you could fit an operating system in just over 3K, but now 3K won’t buy you an emoticon. mad So switching to larger packets would make sense, in that the internet wouldn’t need to engage in quite as much “Did you get that file I sent you? Howbout now? Howbout now? Now?” that it does, er, now.

Since these new features are barely worth the proposed $300 million slice of government pork, the article resorts to slinging FUD on the current internet, boasting that:

The Internet will continue to face new challenges as applications require guaranteed transmissions — not the “best effort” approach that works better for e-mail and other tasks with less time sensitivity.

The one problem here being that e-mail already works like this. As an e-mail passes through each server from its origin to destination, each server gets a confirmation from the recipient when the email is transferred successfully, and the transmitting server will keep trying to send the email until it gets this confirmation. If not, it tells the sender that it couldn’t do it. Simple, elegant — almost as if it was planned out beforehand.

Jonathan Zittrain, a law professor affiliated with Oxford and Harvard universities, shows his ignorance of the internet’s architecture, adding “The network is now mission critical for too many people, when in the (early days) it was just experimental,” forgetting that the internet was a product of the cold war, and as such was designed to be a network between cities such that in the event of a nuclear strike, it would survive even if the cities didn’t. The planning was, if Houston is destroyed in a nuclear attack, the connection between New York and Los Angeles will just re-route itself through Denver. Is this not mission critical enough?

Another one of the reasons for creating a new internet as the dwindling amount of available addresses. In the writer’s futuristic utopia, even the humble toaster will have an internet connection. For some reason. So how can all these new devices have an address when the current system only allows for 999,999,999,999 devices? I mean a trillion devices really isn’t all that many when you think about it. Unless you think, for example, that one trillion seconds ago – 31,688 years – Neanderthals stalked the plains of Europe.

Even taking into account the vastness of one trillion unique addresses, the internet already has a way to handle that: subordinate networks, or subnets. The address range from 192.168.0.0 to 192.168.X.X is reserved for these virtual “sub-internets”. Okay, so maybe there isn’t room for one trillion devices now that the 192.168 range is taken up. sad But in exchange, every network device you add to the internet is addressed inside its own sub-internet, with its own pool of addresses. So now you and everyone else has room for all the toasters, blenders, coffee makers and air conditioners you want to put online; they’re all on our own subnets created inside the internet, like each layer inside a russian nested doll.

These paper tiger problems with the current system and vague solutions offered by the next are what get to the heart of the matter. What exactly are the compelling reasons for this new internet that the current one doesn’t already have? I mean, apart from the wiretapping?

I’ll leave the final quote from Dipankar Raychaudhuri, a Rutgers University professor who is overseeing three clean-slate projects:

“It’s sort of a miracle that it continues to work well today.”

And yet it does. It was a miraculous piece of networking architecture that has withstood an exponential growth in traffic and scale and still manages to handle a load that was inconceivable when the internet’s communication protocol was first introduced. I have a hard time believing a committee of vested interests can come anywhere close to achieving that level of sophisticated programming, and apparently at the National Science Foundation, neither can they: “GENI could start by 2010 and take about five years to complete. Once operational, it should have a decade-long lifespan.” What was that quote about running just to stay in one place?

It sounds trite, but hadn’t these engineers ever heard “if it ain’t broke, don’t fix it”?

And one last thing. With all the spam, phishing, denial-of-service attacks and so on, maybe the problem isn’t the internet itself, but rather the people on it.

Comments

3 Responses to “The Internet: Let’s call the whole thing off”

  1. Myles says:

    I am so sorry I followed that last link. -M.

  2. Anonymous Coward says:

    999,999,999,999 I.P. addresses? What about IPv6? 2 to the power of 128 (2^128) addresses = about 340 billion billion billion billion.

  3. Tim says:

    Yes, that is a lot of devices. In fact, that’s 52 billion billion billion devices for every man, woman and child on earth. While IPv6 is forward-looking, I can’t conceive of the combined effort of the entire human race even coming close to using that many IPs. (Maybe the current IP system could be modified to accept values from 0-F, rather than 0-9. That would open up a few more addresses!)

    While the potential use of IPv6 is sound, it’s all the other invasive security measures of the new systems that bother me. Too much is lost for too little gain.