In honor of Wikipedia’s 15th anniversary last month, we sat down with Ward Cunningham, staff engineer at New Relic—and creator of the very first wiki—to discuss the past, present, and future of his invention and get his thoughts on the importance of Wikipedia.
New Relic: What was the problem you were trying to solve when you created the wiki?
Ward: About 20 years ago, as computers were becoming more powerful and more widespread, I was trying to change the way people wrote computer programs to take advantage of that, via object-oriented programming, among other methods.
Eventually, we latched onto this idea that had been developed by Professor Christopher Alexander, an architect at U.C. Berkeley, and his colleagues, called Patterns. The theory was that before there were architects, people just built structures, and they actually did a better job. Old cities are much more interesting than modern cities, for example, in the sense of being alive. Well, we wanted programs that are alive.
Patterns are things you can see in the city, and Alexander said those patterns form a language just like written language. And while we can study language, when it comes to talking, we don’t have to stop and say, “Let’s construct a sentence. I’ll probably need a subject and an object and a verb in there.” It just pours out.
Similarly, when builders were ready to create, they just spewed out designs. So he asked, “Well, where did that ability come from?” It came from building a house with your father, and with his father and so forth. It was just stuff you did. And as a human you learned it just like you learned to speak a language.
So they set out to develop a form of writing about how cities were assembled. And we said, “We can take that same form, this pattern form, and apply it to object-oriented programs.” He was looking at things present in the city, and we were looking at things present in software.
New Relic: So how did you get from there to the wiki?
Ward: Well, patterns weren’t anything anybody ordered; my boss hadn’t asked for that. But we thought it was a way forward, so we assembled an email list of 500 people and put together the first PLOP (Pattern Languages of Programs) conference in 1994. About 100 people showed up, including graduate students from the University of Illinois, where they were building Mosaic, the first graphical Web browser. One of the students—Brian Foote—very generously showed it to me, told me how to do it, and suggested that the Web’s version of hypertext would be the right way to collect this body of work about how we’re going to write programs. And he told me that I should do it.
I don’t know why they didn’t just do it themselves. But I liked the idea. I started by asking people to send me patterns as text files, and I’d convert them to HTML. But that was boring. So I made a website where they could submit their patterns through the Web, and then I would process them into HTML and show them what they got. And that was wiki.
New Relic: So what exactly is a wiki, anyway?
Ward: It’s basically a way of writing where you’re reading. On the Web before that, you would read something in one place but if you wanted to write more, you would have to go through a completely different mechanism. You couldn’t author through the Web before that.
New Relic: Would you consider Wikipedia to be the best-known application of the wiki?
Ward: It’s certainly the one my neighbors know!
New Relic: How did your work on the wiki lead to the formation of Wikipedia?
Ward: They picked up on wiki after it had been in existence for five years. Ben Kovitz was one of my regular contributors; he’s a computer programmer. He knew Larry Sanger—one of Wikipedia’s co-founders, who along with Jimmy Wales was trying to put together this thing called Newpedia. But after working on it for a year, they had gotten only about five articles or something on it … the pace was too slow.
So Ben suggested to Larry that they put together a wiki for authoring of the articles, and then experts would come and proofread them. But as the community formed around the wiki it became a very tight loop—it’s write, publish, and then review—so reviewing ended up happening after you publish. That makes it a lot easier for different people to apply their various expertises. It’s basically an open source project where strangers meet and create value together.
New Relic: How involved are you in Wikipedia today?
Ward: I’m an advisor. I’ve met all their executive directors. I see them struggling a bit to make good on their promises to their contributors. I follow the academic research it has spawned. I’m a fan but not an editor.
New Relic: What’s next for the wiki?
Ward: I’m working on developing the new wiki. After 20 years of watching the wiki develop and being disappointed in some regards, I think I know what the answer is: what I call a Federated Wiki.
I’m four years into that project. The first year with support from Nike, where I was a Code for a Better World Fellow, and three years with a community of developers around the world that have tried all kind of things with it.
To participate, you create your own server. Think of it like a blog: If you want to have a blog, you put one up on a server somewhere. Here, if you want to participate in the federation of wikis, you put up a wiki. But they all share in the same way using the Creative Commons and so forth. We’re up to more than 800 sites now; most of them added in the last year.
The Federated Wiki allows users to share their data but still own it. No one can modify it and say it’s what you created. Instead, I write to my wiki, and you write to your wiki. Anything you write, I’m welcome to take via Creative Commons with built-in attribution. But when I revise it, I put it on my wiki, not back on your wiki. So somebody could read my page and say, “Wow, Ward, you’re really doing great stuff here.” And then they look and see that I really wrote only one paragraph and it all came from you. It’s kind of a collective ownership because I’m giving it away—and the nice thing about digital stuff is that I can give it away and still own it.
So here is the big win: The people who write for Wikipedia spend a tremendous amount of energy correcting abuses. Federated Wiki solves the problem because when I put something on my wiki server, you’re free to use it but you can’t screw with my original. And if your version is better, you’ll get the eyeballs, but then I’m free to take your version back. Pretty soon, we’re advancing big problems in the world.