Weaving a better Web

Experts offer their take on the future of the 'Net

The Internet goes Interplanetary

In 1999, Vint Cerf pitched the idea of an interplanetary network that would extend the Internet's reach beyond Earth. Even though it came from the man who had co-developed the Internet in 1973 and was at that time a spokesman for MCI, the idea seemed far-fetched to many in the industry.

Almost a decade later, Cerf, who is now chief Internet evangelist at Google, has not wavered in his belief in the idea. "By the end of this decade, we'll have a two-planet Internet in place. We'll have software on orbiters that allow new protocols to make the Internet work across the solar system. This is a very exciting prospect," Cerf says.

You need only look at the billions of dollars the National Aeronautics and Space Administration has spent in the past few years to explore Mars to see that Cerf's vision is creeping closer to reality. "It's only been eight years for this interplanetary idea to evolve. When you think how long it took the Internet - 10 years just to get the basics in place - this is moving quickly," he says.

Cerf is just one of many luminaries placing bets on what the future holds for the Internet and the Web. And although other predictions are not as out of this world as Cerf's, they offer a glimpse into what we can expect from the Web in the next few years.

The Web as a platform

Tim O'Reilly, who co-developed the idea of Web 2.0, a vision for the next generation of software, says the biggest change we'll encounter is a move away from traditional network architectures to having "the Web as a platform."

"The first thing we need to realize is that this isn't just a software revolution," O'Reilly says. "It's a revolution in the way we capture and coordinate data."

O'Reilly says there's a misconception in the industry that today's databases will be sufficient to handle all the user- and machine-generated content - petabytes per year for some companies. "Google's not managing their data using [traditional databases]. They're developing very new types of tools," he says.

"Other companies must develop competencies in managing, visualizing and extracting information from massive amounts of data to take the Web to the next level," O'Reilly says.

One thing to expect is a whole new crop of applications that will be generated from data. "Applications will now be built on stacks of data services," O'Reilly says. Map applications are a crude example of what's to come. They are formed as they draw content from a bunch of different providers, he says.

O'Reilly says the key to succeeding in this data-focused world is to find new ways to harvest the massive amounts of data that will be generated. For instance, he says cell phone companies soon will figure out a way to use the vast amount of data they hold. "They know who we've called, but they don't know how to harvest that information to create a business. There will eventually be people who realize how to mine that data in the same way Google does Web searches," he says.

One obstacle to O'Reilly's vision is a fear in the industry about the dangers of openness and data sharing. "You have to figure out what pieces you get to control and own to gain a competitive advantage, and what you can give away," he says.

The benefits of crowdsourcing

Scott Anthony, a colleague of industry guru Clayton Christensen and managing director of Innosight in Watertown, Mass., says social networking, which encourages companies and individuals to collaborate via technology, is helping to drive openness.

"Social networking enables companies and individuals to do things they couldn't do in the past. No individual or group has a license on good ideas," Anthony says. One example is a concept called "crowdsourcing," in which there's open innovation and collaboration among many parties, he adds.

Anthony says there has to be a shift from a culture of "invented here" to "found elsewhere." "People need to see that borrowing from the outside or grabbing from somewhere else is a good thing. The notion that you want to solve problems on your own - that's not the way innovation works. Companies have to change their cultures to move forward," he says.

Federated identity is key

Dave Passmore, research director at the Burton Group says that federation is critical to data security. "We're going to see more social groups where companies get employees and customers

to interact in a peer-to-peer fashion. There will be a lot more socialization via the Internet," he says. For social networking to succeed, there will have to be an increased focus on identity management, Passmore says. "We still don't have a good way of identifying individuals on the Internet," he says. "This can't just be authentication; it has to be some authority that can vouch for you in a scalable way. Today, every time you join a group, you have to enter a ton of user data," he says.

Passmore says authority, coupled with a concept called "federation," could ease security fears. "Federation allows you to carry your credibility from one site to another," he says. For instance, a user could develop a solid reputation on eBay and carry that over to another site without having to reestablish his identity. "This doesn't mean that anonymous participation goes away, but for the activities where you need to know that there's a live human being on the Web, there will be a system in place to protect you," he says.

At long last, IPv6

Tim Winters, software managing engineer with the University of New Hampshire Interoperability Lab,

predicts that the move to secure granular bits of data

will finally mean the emergence of IPv6.

The government is aiding the push to IPv6 by mandating that civilian and defense agencies support the protocol in their equipment by mid-2008. Organizations might be surprised, however, to find that a lot of applications already support the protocol, Winters says.

The IPv6 movement has had a tough go over the past few years. Experts have warned that the IPv4 address space is running low, but network address translation (NAT) techniques have helped companies extend their use. But now, security concerns are starting to shed light on the benefits of IPv6, such as end-to-end security.

IPv6 also will help with the influx of wireless networking and streaming video. "You could go end to end without going through a NAT box, which will improve quality and speed up the delivery. You can also secure the video stream so it can't be tapped into," Winters says.

Microsoft Vista is a boon because it has native support for IPv6, he says, and other application vendors are testing IPv6. "In the next couple of years you'll see IPv6 become more prominent and widely deployed," he says.

Data security takes center stage

Stan Quintana, managed security services vice president at AT&T in Bedminster, N.J., agrees with Passmore that federation is critical to securing data moving across intranets and extranets. Federation should be linked to a company's network and data management policies, he adds. A company that acquires five other companies can meld the policies and create a federation that lets them share data safely.

In addition to federation, the next generation of the Web will include a shifting from perimeter-based security to data-centric security, Quintana says. "We're going to put security in place at a granular level to protect information itself," he says.

The approach will be multipronged. "I have to encode it. I have to ensure the access to that information - the keys - goes to the right people and is well managed. I have to make sure data is only useable on a need-to-know basis," Quintana says. Other factors include classifying data and making sure it can be distributed without being modified in flight. "Once you have all these steps in place, you'll be able to open up your environment because your data is safe," he says.

In the future, all savvy companies will have a sophisticated digital-rights platform that will let them define and classify their information, as well as map it to appropriate users, Quintana says.

He also believes there will be universal authentication. Like Passmore, Quintana thinks that single sign-on, which has been promised for some time, will happen, and a central authority will hold a user's credentials. These credentials will function as authorization for the corporate infrastructure and commercial entities. He says biometrics, such as an iris imprint, will be used in conjunction with other identifying factors.

Quintana also predicts that security will become a commodity and will be built into other services. As infrastructure, such as servers and storage, grows exponentially, IT managers will give up handling all security problems themselves, he says. "I don't think people realize today how hard it is to scale customer-premises-based security," he says. "This is going to take everyone by surprise."

He believes companies will realize the cost of internal security management in terms of risk, personnel and software, and subscribe to software-as-a-service in the core of the network. "The magnitude of very sophisticated exploits - not just free-form viruses - is growing. Crime syndicates are putting in place targeted attacks, spyware and malware," Quintana says. Individual companies will have a hard time keeping up with solutions to these security problems, and will need help, he adds.

Sensor-based networks

Another big area that's ready to take off is sensor-based networking. Johna Till Johnson, founder of Nemertes Research and a Network World columnist, says sensors are the key to the long-promised machine-to-machine communication.

"It's a system-to-system, Web-to-reality interface that doesn't require human mitigation," she says.

Johnson says sensors, such as RFID chips embedded in everyday products, offer infinite possibilities for gathering critical data. For instance, sensors on ocean buoys could give citizens advance warning of storms. "This real-time communication can be tracked against online databases to create powerful information," she says.

Johnson admits there could be legal problems regarding privacy. For instance, if a car manufacturer puts sensors in its cars that help locate a driver in an emergency, there is a risk that the information, gathered in a database, could be used in other ways. "If databases were merged, they might give levels of data that are unacceptable. We need rules and regulations to guard against that," she says.

Gittlen is a technology writer in Northboro, Mass. She can be reached at sgittlen@charter.net.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2007 IDG Communications, Inc.

IT Salary Survey 2021: The results are in