In 1993, around the time that everyone decided the Internet was all text - email, file transfers, chat, and Usenet - along came the Web. In 1998, around the time that everyone figured the Internet was all about the Web, ecommerce, and Big Servers, along came person-to-person chat services like ICQ, peer-to-peer file-sharing like Napster and Gnutella, and distributed computing experiments like SETI@h ome and the Uni ted Devices cancer research project.
All of these things use the Internet; none of them have anything to do with the Web once you're past the initial find-the-software-and-download-it-stage.
At O'Reilly's peer-to-peer networking conference the other week ( here), you got the chance to see that these things are not non-recurring phenomena (in the phrase made famous by Hollywood executives at a loss to explain unexpected hits). They are beginning to use the Internet as the unique medium it is. Email is kind of like telephone calls or letters (one-to-one). The Web is sort of like publishing or broadcasting (many-to-one).
But what's unique about the Internet is its potential to connect anyone-to-anyone and many-to-many without the intervention of a central authority. Usenet (and bulletin board systems before it) was the first hint of this, as it enabled unmediated, public, social interaction at a distance - something never available to most people before. (I say most people, because when I wrote my first book, I suddenly realized that book authors, as well as other publicly creative people such as musicians and film directors have always had ongoing, public conversations with each other, but the barriers to entry were always fairly high.)
Because of the ferocious copyright battles that have surrounded Napster and Gnutella, everyone assumes that peer-to-peer networking is all about "stealing" music, a notion that Sun's technology evangelist, Simon Phipps, speaking at the conference, called the "baby duck phenomenon." You know, the baby duck hatches, and whatever moves first in its field of vision it thinks is Mother. Well, Phipps said, peer-to-peer networking hatched and the first thing it saw was an MP3 player so it thought it was a copyright- violation mechanism.
But in fact what we're doing, said Clay Shirky, a writer and teacher at NYU, is inventing a new operating system for the network. In 1998, it was possible to imagine that a Web browser was the only software you were ever going to need. That was why Microsoft and Netscape battled so intensively and how the US Department of Justice got involved, even though: people thought the browser was going to be the platform for everything in the future.
It's amusing to note that net.wars, the book for which this column is named, was frequently assumed to be about the browser wars back then; now, people ask if it's about cyberwar. It's about neither, but the border wars between cyberspace and real life, like this column.
Back in 1993, if you got onto the Internet, you had a TCP/IP stack/dialler such as the geekily named KA9Q, an FTP client, a mail reader, a Usenet reader, and a Telnet client. Once graphical Web browsers became widespread unifying interfaces, increasingly you needed fewer and fewer bits of software. But cut forward to 2001, and what do I run every day? Netscape, to be sure, and Ameol for email and Usenet because I'm a pipe-smoking CIX curmudgeon, but also ICQ, AIM, LimeWire (for Gnutella), and UD Agent.
So you can see Shirky's point: just as the operating system makes it possible for your computer to connect to your printer without requiring each application to reinvent the driver, so it ought to be possible for applications across the Internet to share standards and protocols so that someone sharing files via KaZaA doesn't have to run two pieces of software to also connect to Gnutella. Jabber is one of the first steps toward this world, as it lets you use any of the major instant messaging programs from a single interface. France Télécom took an equity stake in Jabber in July.
XML, which got a lot of ink a few years ago and doesn't get talked about as much now that people are really beginning to use it, is a key element. Making the Web machine-readable and data interchangeable between programs using the same tagging system goes a long way toward making resources usable wherever they're found. One consequence will be longer and more protracted intellectual property battles, as people create new Web services that extract data from multiple locations around the Net (including the Web) and put them together in new ways to create information that wasn't available before. Quite often, those uses will be quite different than the original owners of the data intended
So by the end of a week listening to people talk about P2P, I wound up with a dim glimpse of the networked world in a few years' time: one where the Internet is restored to its original decentralized design.
That's the one that can withstand a bomb dropping on it and the one of which John Gilmore once said: "The Net perceives censorship as damage, and routes around it." In a lot of ways, it's a very comforting vision.
* IN response to last week's column, reader Chris Hall wrote to remind me that DeCSS decodes DVD .vob files, which have been encoded according to the Content Scambling System or CSS, but that although it allows software "rippers" to copy DVD files to hard drives in a decoded form, it is not the region coding, which instead is implemented in the drive itself as well as on the disc. µ
Previous Columns
S ave the Cookie
n et.wars: Digital rights and the new era of world terrorism
Wendy M. Grossman, whose Web site is pelicancross ing.net, is author of From Anarchy to Power: the Net Comes of Age (NYU Press, 2001), net.wars (NYU Press, 1998), and the Daily Telegraph A-Z Guide to the Internet (Macmillan, 2001). She can be reached at this email address.