Ultranet, the next network

A bleeding-edge federation of startup companies strives to reinvent the next-generation network

The Internet today is a different creature from what we remember in the 1980s. At that time, the dominant domain was .edu; now the dominant domain is .com. Traffic on the Net has increased by a factor of 200. Radical changes like these create new opportunities and new challenges. The Internet continues its growth, and each stage presents new opportunites and new pitfalls for the companies involved.

Stage one: the rise of the Internet

  • Defining technologies: Internet Protocol, SNMP, HTML
  • Business model: giving away information and software for free
  • Winners: creative small companies -- Netscape, Sun Microsystems (JavaSoft), Yahoo
  • Ecological metaphor: new ecosystem

During the early days of the Internet, the only way to transfer value through the network was to give things away for free. The companies that emerged as successful at this time were those explosively creating mindshare through the distribution of free software -- in particular, software that, by its nature, represented the architecture of the Internet. Like hardy, new plants that appear after a highly destructive forest fire to define the biological substrate upon which all following species depend, these pioneering companies exhuberantly charged into this new fertile network territory.

Although there were many winners in this pioneering era of the Internet, I will pick three exemplary cases: Yahoo, Netscape Communications, and First Person (the division of Sun that eventually became JavaSoft). I refer to them all as "small companies," since Yahoo started as two people, Netscape as a half-dozen, and First Person as a few dozen.

Leveraging the successes of Mosaic, the engineers at Netscape jumped ahead to claim a dominant position in the browser space by giving away the Navigator software. Nipping closely at Netscape's heels was Sun, which contributed Java to the architecture of the Internet. The dynamic duo of Filo and Yang created the Yahoo Web index, which remains among the Web's most popular destinations. This contribution to the Web architecture is paid for by advertisers and sponsors, and represents another successful business model for this stage.

Early arrivers to the Internet essentially were driven by their interest in investors' money and were giving away their products and technologies for free. The idea that customers could pay nothing for a company's products and investors would line up at the door to invest in them is a strange one. But grabbing investment cash with both hands and tossing away value is not a sustainable practice for companies. There are still those who don't understand the business model of open software: publish the interfaces and compete on the implementation. Yet at the time, this practice demonstrated a powerful truth: The contributors of innovation are well rewarded in the marketplace -- a truth not applicable to all markets at all times.

Stage two: Growth of intranets

  • Defining technologies: firewalls, groupware, Web/SQL connectivity
  • Business model: giving away information and software for free (but only to members of my own company)
  • Winners: stable, medium-sized system integrators -- Baan, SAP, Anderson Consulting
  • Ecological metaphor: the segmentation of the biosphere into biomes

The discovery of a sustainable business around the distribution of free information and software caused another rush of excitement. The idea of allowing members of a company to give each other free information and software through an "intranet" spurred an explosion of development activity. Ultimately, though very lucrative for system integrators and companies who serve the IS departments of the Fortune 1000, this phase did not represent a major shift in the available technology base. It merely consisted of each of the major company types adapting the existing technology in a way that suited its unique needs.

Stage three: extranets

  • Defining technologies: "Crossware"
  • Business model: Giving away information and software for free (but only to members of partner companies)
  • Winners: Big Large companies -- Netscape, Microsoft, Sun
  • Ecological Metaphor: Eutrophication, a lake turns swampy

Once a large number of corporate networks were running on open Internet Protocol it was a simple breakthrough to realize that you could interconnect the intranets, thus creating an inter-intra-net. Always willing to create new words for what continues to be essentially a business repackaging of stage-one technology, the words "extranet" and "crossware" came into being. The essential value of these is to enable access to your partner company's Web pages (and hence database, in many cases) from within your own company's internal Web page. It's unclear that there will be any big winners in this arena -- it's just too early to tell.

But I will characterize the Extranet stage as the last stage in the evolution of the idea of giving away free information and software, an idea that gradually may be reaching the limits of its value. How many ways can you repackage the idea of giving away information for free? One indicator that the ideas are starting to play out is that big, slow-footed companies now are starting to reap the benefits of the stabilization. In nature, this process is called eutrophication, and can be defined as the process by which nutrients are added to a system. The culmination of eutrophication on a lake, for example, is the nutrient-rich swamp, which suggests that the growth remaining in this ecosystem is declining.

Stage four: the ultranet

  • Defining technologies: Java, IPv.6, SET, OODBMS, XML
  • Business model: value exchange of goods and services (open, complex market economy)
  • Winners: creative companies of all sizes -- to be determined

I'm going to indulge in some gratuitous, tongue-in-cheek word invention here. I'm going to invent a name for a network (the same exact network as in stages one through three, actually) that's so magnificently superlative that it will put a stop to all of the Internet-intranet-extranet naming games. This name is the name to end all names, and characterizes the mother of all networks: the ultranet. What are the defining characteristics of the fourth stage of the Internet? This phase is foreshadowed by the invention of IPv.6, Internet Protocol version six. The very protocol that lies at the heart of stages one through three is being reinvented in order to provide more address space. How much more space? Up to 133 million addresses per square meter of the earth's surface! This suggests an Internet with orders of magnitude of more people and devices connected. How will we accommodate browsing information in these new ways?

The XML standard, as released by the World Wide Web Consortium (W3C), holds the promise of deconstructing the HTML Web through well-defined interfaces. This means that your alphanumeric pager can grab a small piece of a news page -- a headline or a stock quote, for example. This form of structured interface creates the possibility of Web automation. Web automation allows collections of databases and people to chain functionality and aggregate information over multiple "hops," instead of using a strictly client-server model that today's Web holds. But going beyond simple structured interfaces, XML defines extensibility through programmer-defined mark-up tags. If a browser environment encounters an unfamiliar tag, it downloads an executable that allows for rendering of the content. Ultimately, structure and extensibility enable an object-oriented approach to take hold on the Web. The structure can be seen as an Interface Definition Language (IDL), and the extensibility can be seen as a direct way to marshall an object from a remote server.

So now that XML has created an object Web, what does this innovation do to the fundamental structure of the Internet? This is where Object databases enter into the scene. With a world of objects and interfaces (IDL), there comes a need for a dynamically populated repository. This, of course, is differentiated from storing HTML in a simple flat file system, and much more powerful. With objects, the information that is stored and delivered is not just data, but executable code. I'll get into why this is interesting in a moment. But first, I'd like to mention a class of objects that suddenly appears once you have a repository of XML data objects. What tends to appear are "people objects"; this is part of a trend that is already anticipated by the OPS systems, a universal system for carrying user profile information.

Don't you get tired of typing all of your personal information into every single Web site you visit? With the appearance of people objects in these object repositories, we have a way of representing not just data assets but people as well. Do you recall I mentioned that these people objects have executable code in them? Well, what this suggests is that if they have network mobility code, they are "agents," and if they have a user interface, they are "avatars." Now that we have a population of people objects and asset objects, it's key to have some way to assess who is interested in what, so we can allow collaborative filtering to increase the signal-to-noise ratio of the Internet. One company that is showing a lot of leadership in these areas is Affinicast, a small startup based in the South of Market Street area in San Francisco. Another beautiful thing about all of these objects is that you can aggregate all sorts of data around the objects--data that can be publicly addressed and manipulated.

The notion that some of these objects are devices, some of these objects are data assets, and some of these objects are people opens the possibility that new classes of applications and services can become available. But the driving force of the services economy will be the advent of electronic commerce technologies such as the Java Card, Secure Electronic Transaction (SET), and the Java Electronic Commerce Framework (JECF). The security involved enables an authenticated secure channel to be opened between any arbitrary number of points on the Internet. This means that "virtual intranets" will spring up and dissipate at a rate that makes Silicon Valley corporate mergers seem glacial. Securely delivering economic value (money) to anyone who delivers value through the network will be the driving force that transforms the collection of Inter, intra, and extra nets into the ultranet.

This form of economic freedom can be paralleled only in the lush tropical jungles of the Brazilian Amazon. This ecosystem is the most fecund biosystem in the world. The biodiversity of this ecosystem is reflected in hues, chirps, ululations, and the dense tapestry of light, sound, and smell. And who is king of this jungle? If you said "lion," give yourself a B for understanding the popular myths of our culture. Just to give you a sense of what kinds of companies might win in a "jungle economy," take note that the lowly rainforest ant has four times the biomass of all of the mammals combined. Watch for self-organizing communities of netizens to prevail in this space.

Miko Matsumura has a Masters degree in Neuroscience from Yale University and a BS in Psychology from the University of Michigan. Before becoming the Java Evangelist for Sun Microsystems, he worked at HotWired as the director of Research and Development, at the WELL for Woodstock '94, and at the Branson School. He can now be found at Miko.com. Miko's first computer was an Atari 400 with 16 kilobytes of RAM and a cassette tape recorder, which he got at the age of 12. He has been pondering questions about human and machine behavior ever since. He holds a first-degree black belt in Shotokan karate.

Learn more about this topic