Great X-pectations: Executable, extended--new Internet paradigm could change the user experience and connect us with the real world. (Business of Technology).

When Carl Howe, principal analyst at Forrester Research, issued his report on the X-Internet just over a year ago that proclaimed the Web was fading fast, he had some experts virtually foaming at the mouth. Critics dismissed the idea as nothing more than a buzzword, saying that with the Web in a constant state of evolution, it was foolish to announce the end of one era and the beginning of another.

But a number of startups and some very high-profile old timers, such as Apple and Sun Microsystems, have taken the concept seriously. One year later, there are some intriguing developments that may just bring us closer to an Internet that enables greater interactivity with less hassle, on devices that go far beyond the desktop PC or even consumer handhelds.

"What the Web did was make us very tolerant of bad user interactions," said Howe. "We got used to things that take 10 to 15 minutes to respond. Basically, it made people wait for all the things they want to do on the Net. And weren't these machines supposed to make us more efficient? The Web is great for things you read, but if I'm actually trying to do a transaction or have a media experience, it's dreadful."

The real key to this X-Internet concept, he emphasized, "is to eliminate the world wide wait."

What constitutes X-Internet? On the one hand, it's, in Forrester's words, "executable" and about software and user interface. It moves code more efficiently from servers to user PCs and other devices to reduce latency and enhance greater interactivity online. On the other hand, it's "extended" and about hardware--those things, be they refrigerators or assembly lines, that touch the real world.

X-Internet is not to be confused with the Semantic Web, which will give the Web the ability to make logical associations among words and concepts. Although X-Internet and the Semantic Web are being developed on separate tracks, their goals are compatible and will likely converge as research on both fronts advances.

Beyond Dumb Terminals

We all know what it's like to try to execute a transaction online, whether it's buying a book or making an airline reservation--lots of clicks on a series of Web pages that may take ages to download. And if you need to return to a page? Click on back.

It's the dumb terminal syndrome. All the information resides on the server and needs to go back and forth through your Internet connection. Several new companies are in the trenches developing tools to add smarts to your terminal. Altio, for instance, has a platform that delivers rich Internet applications, according to founder and CEO David Levett.

"We basically have created an incredibly powerful network where in theory, anything can be connected to anything else," he said. "We can run applications that look and work like Outlook or Quickbooks, and run very fast. It downloads an application temporarily with just the features you need to accomplish the task. There's no installation required. It may be about 200K or a bit more than one page in terms of downloading, then it can be cached. Everything is instant and the end-user result is that there's much better functionality. You don't have to keep going to the server and back again."

The footprint, said Levitt, is small, using only one-thousandth of the bandwidth that would ordinarily be required because it's encoded in XML and sent across compressed, with only the modules necessary for a given task. "We're doing all the hard work behind the scenes like an operating system, all within code," he said, "So that 200K is doing an awful lot of work."

Curl Corporation, too, is developing a more dynamic approach to executable Internet. John Capobianco, executive vice president of strategic planning, said, "It's not just that I want formatted content; I want data sent to me so I can decide how to use it."

To that end, Curl has created a platform that shifts the emphasis from client/server to what it calls client/Web. According to Capobianco, Curl applications are a set of development tools that combine scripting language, text processing and object-oriented programming. The application code sits at the server side. When the user issues a URL call over the Internet, it goes to the server, which notes the location and sends the Curl application back to the user. This application is compressed and executes quickly. The application, he explains, runs on the client side and every time the user wants additional information, the Curl application requests that the server send the data. "It's a fat client without the fat," said Capobianco.

Then there's Macromedia and its MX line. Says vice president of strategy Peter O'Kelly, "With MX, things are integrated. An MX-enabled site will allow transactions to be performed on one page that is automatically refreshed or updated as the transaction takes place. "Traditionally, you have distinct tools like email. But it's not the way people work or want to work. Our vision with MX was to take the best of Web applications and extend them further by taking advantage of the full capabilities of the devices. We think about the task. The tools should be so natural and integrated that you're not even thinking about them," he explained.

The business case for these applications, said Altio's Levett, is straightforward. "The application requires no downloads or plug-ins, it runs through firewalls and can deploy to people in your extranet without having to go out with a CD and do support for them, and it provides diagnostic information for people using the product."

While Altio is working with companies that want to use its products in an internal enterprise environment, Curl and Macromedia are looking at potential consumer uses as well.

These kinds of applications provide a strong value proposition to Larry Hawes, senior advisor at The Delphi Group. "These [new tools] allow you to decrease development time and get applications to market faster," he said. "The other, more important, issue is user productivity. The idea of interactivity over the Web now is limited to a controlled series of HTML pages, but these new tools make the user interface more responsive and intuitive," he said.

Pick a Device, Any Device

If the idea of an executable Internet sounds new and exciting, the idea of an extended Internet has actually been around awhile, which Howe, of course, concedes. Telematics is a kind of extended Internet application. So are instant messaging, wireless sensors and plug-and-play. And as innovations like microelectromechanical systems (MEMS), Wi-Fi, ultra wideband and others mature, the possibilities expand. Apple, for example, has just released Rondezvous, which creates instant wireless networks of computers and smart devices.

Sun's Project JXTA is another notable example. It enables services to connect. Essentially a peer-to-peer technology, it's an open source platform focused on collective processing.

"JXTA attempts to define a set of protocols for peer-to-peer computing that allows for collaboration across multiple computing environments and multiple physical networks," said Matt Reid, business development group manager for Sun. "We have to be able to include not just TCP/IP networking, but things like Bluetooth, infrared and cell phone networks. This plethora of devices and physical networks creates a huge amount of complexity for the developer. The core idea is to create a virtual network in which the complexity is obstructed out. The developer treats the network as one network and JXTA handles all of the underlying complexities."

The result, said Reid, is an environment in which multiple devices can be leveraged so that the technology functions more like a person would work. That alleviates the user from having to adjust to the way the technology works. So with a personal peer group of matched devices, you could update a phone number on a cell phone and that new number would automatically propagate on all the other devices in the network: the desktop computer, the PDA, the laptop. From an enterprise standpoint, notes Reid, it could mean better use of computer resources that sit at the edge of the network.

Getting computers to sense the real world is a challenge for the Auto-ID Center, a research project of Massachusetts Institute of Technology funded by about 65 industry sponsors. Researchers believe that by putting microcomputers into every manmade object in the world, computers could, in a manner of speaking, sense the real world.

"It solves the problem of identification," said director Kevin Ashton. "What we're trying to do is kick off the next 50 years of computing. The purpose of the Auto-ID center is to get machines to perceive things on their own in the real world." Using an electronic product code embedded into products allows companies to pinpoint in real time, the product's location so an action can be taken. The Internet creates the network where data is stored so that it doesn't have to be stored on the actual object. This, he notes, reduces the cost and makes it possible to apply on the wide scale necessary to address supply chain issues.

As appealing as these technologies are, some experts have reservations. Gene Monacelli, national telecommunications leader, management solutions and services of Deloitte & Touche's TMT Group, is concerned about the privacy and security issues for these types of applications. "That means you're running devices over different types of networks, interfacing between the network and the device wirelessly and going through someone else's wired exchange," he said. "All this has to have security built-in. A lot of folks are going to say they're not interested in putting a lot of money into the necessary security to make all this workable and secure."

Monacelli also believes that people need to be very clear about the business benefits of using these technologies and how they'll be deployed.

The timeline can be several years out for these technologies. Howe doesn't see the extended Internet really touching consumers in a major way until 2007, due to the expense. And, he suggested that enterprises cherry pick technologies that would decrease costs or increase revenues.

Auto-ID Center's Ashton, for one, is optimistic about the potential.

"This is what the Internet was really invented for," said Ashton. "It's a foundation stone to enable machines to perceive the real world. That's what 21st-century computing looks like."

RELATED ARTICLE: Tip: Putting the X-Internet to work

Internet is going to be an evolutionary process. In the early stages, companies need to decide where to deploy their resources for the greatest benefits.

* Don't try to pick X-Internet winners this early. Instead, invest your time and money in X-Internet experiments that yield business value today.

* Start using wireless networks so that when you want to roll out an extended Internet application like production or supply chain monitoring, the network will already be in place.

* Identify high-value extended opportunities- places within the corporation where sensing the physical world in business would decrease costs or increase revenues.

* Allocate training dollars to learning the applications.

* Ask where your users need a better experience.

* Ask where you need more scale and have less need for bandwidth.

* Ask how you can put more intelligence near your users.

Source: Forrester Research

Naresh Lakhanpal is director of the Product Innovation Group for Deloitte & Touche Technology, Media & Telecommunications (TMT) Group

www.deloitte.com

www.altio.com

www.macromedia.com

www.curl.com

www.sun.com