"The Net interprets censorship as damage and routes around it."
-- John Gilmore

Who Owns Information Technology?

The Internet allows us to share all kinds of information around the world, almost instantly. It's truly a modern miracle - a thing of beauty, both for it's complexity, and it's simplicity. My training and background allow me to understand the inner workings of the Internet, and it remains a thing of beauty. It's a network that operates almost completely without any central organizing structure. It's a system where the overall responsiblity for operation is parcelled out to all of the individual pieces.

When I got started on the Internet back in 1988, it was relatively small compared to today. It consisted almost entirley of universities, research facilites, and technology companies. There was no web, and so it looked entirely different.

Then, as now, email was probably the most important Internet service. From the outside, email is terribly useful. From the inside, email is a thing of beauty itself, much like the Internet. No central authority is required for email to operate. It can tolerate outtages, and successfully deliver mail.

There were other useful services too. ftp was very useful, even though there were no hyperlinks. Ftp was also free from centralization. If you wanted to run an ftp server on your system, you'd just set it up and tell people about it and they could upload or download to their hearts content.

Usenet news was another terrific service. It was something like the web forums we see today. But also something like email. Because much like email, and the internet itself, Usenet News was a system where news would propogate around the Internet from server to server with no central organizing authority.

I'm sure you see my theme by now. There was no central organizing authority to any of it. Most services would run freely without any intervention from "The Man". They'd even keep on running if "The Man" tried somehow to break things.

When the web came along, it fit well into this model. It's essentially like ftp. No ftp server really depends on other ftp servers. You don't need permission to set one up, you can just do it. The web and ftp aren't exactly the same as email and Usenet News. Email and news were push services that pushed the data to where it needed to go, even routing around problems and continuing to function. The data lives everyehwere on the Internet. The web and ftp are on demand services where users grab data as needed. The data lives on individual servers, and it's subject to control at the whim of whoever runs that server.

Politically, one might try to make analogies that the web and ftp are capitalist: one site designates itself to fill some need, and provides the data for that service. Email and Usenet News are communist (in a pure theoreticaal sense, not the real world corruptible centralized sense): The Internet comes together as a community, with everyone doing their part, to distribute data in a timely and reliable fashion. I don't make this political analogy to introduce bias in either direction, only to illustrate the mechanisms of the underlying technology. Each of these services works exactly as it needs to, for it's purpose. Immediate communications are well-suited towards a communist infrastructure. Other things, like sharing a library of constant or growing information, are better suited to the capitalist infrastructure.

The new Internet has forgotten about the old Internet. These days, instead of designing services for the Internet, we design services for the web. The web is considered ubiquitous, easy, and familiar, so delivering a new information service on top of the web just seems natural. And we've created some popular services. Facebook and twitter are two great examples.

From a technology standpoint, this drastically simplifies the development of the technology. Twitter was written from "off-the-shelf" software packages stuck together like a kid playing with Legos. I could rant about why this is bad, but that's not what this article is about.

This article is about the implications of building information services on top of the web, instead of on top of the internet. Let's take a look at twitter. Twitter is an information sharing service, and a communication service. Looking at the old technology, twitter is more like email and usenet news, where people share information, and that information flies all around the internet, going to whoever it needs to go to. If twitter had been designed 15 or 20 years ago, someone would have sat down and designed a communication protocol by which tweets could be transferred all around the internet from one computer to the next. Then based on that protocol, software developers would distribute servers all around the internet, and those servers would collectively be responsible for distributing the data.

But that's not what happened. Designing protocols is hard. You have to understand things like UDP and TCP and dropped packets and sliding window acknowledgement. Besides, that sort of thing just isn't fashionable these days. So twitter was built on top of the web instead. All you need for that is a database and a webserver. And so twitter was built on top of the web. Twitter runs at one single site, twitter.com.

Facebook, and Myspace, and others like them are all the same. These are information sharing services which are well-suited to the communist model of information distribution. And yet they are all built on the web, which is inherently capitalist. As new things come online, like Google Buzz and Google Wave, it looks like this trend is going to continue.

Even with email, the trend is towards a few huge sites like Gmail and Yahoo that handle everyones mail.

You might wonder why I seem to think this is a bad thing. Well, because we're losing everyting good about a distributed Internet. One issue is cost. Under the Internet model, costs are hidden. Each site pays a fee to participate on the internet, but that cost is considered to be the cost of doing business. The same fee that gives the access to

Security is also at issue. The distributed nature of the Internet, and of any services which are similarly distributed, makes it difficult to hack more than a bits of information at a time. But with these new web-based services, it's one-stop shopping. If someone hacked Twitter, the sky's the limit.

There's also the question of control of information. Twitter and Facebook both implement policies


Tom Fine's Home Send Me Email