Data Model Evolution

Only once has the IT industry gotten it right in predicting the timing of technology spending—and that was because Y2K compliance efforts were scheduled, whether anyone knew it or not, hundreds of years in advance.

With due respect, therefore, for the chance that anyone takes in making any nontrivial IT prediction, eWeek Labs shared a look into the crystal ball with five well-placed prognosticators: Rod Adkins, general manager of IBMs Pervasive Computing Division; Rick Rashid, senior vice president at Microsoft Corp. and head of Microsoft Research Group; and a troika of Sun Microsystems Inc. gurus, Chief Technology Officer Greg Papadopoulos (now lecturing at Massachusetts Institute of Technology), Chief Researcher and Science Office Director John Gage, and Sun Fellow and Vice President James Gosling.

eWeek Labs came to the table with a draft list of candidates for top technologies of next year and beyond, including:

• broadband networks;
• Web appliances, all form factors;
• open-source software;
• wireless network access;
• self-healing systems;
• programmer productivity aids and
• identity management.

Suns Papadopoulos was quick to point out that these categories overlap in ways that arent widely recognized. Wireless, for example, is not just the relatively low-bandwidth solution to meeting mobile needs. "Fixed wireless can deliver bandwidth of 100M bps to any home or office within range of a relay station, significantly reducing the last-mile costs that have held back broadband deployment," he said.

At the same time, Papadopoulos, based in Santa Clara, Calif., agreed that the proliferation of mobile users and applications to serve them is driving a new point of view on system design.

"The Internet is rapidly moving beyond connecting computers and their browsers to connecting people and providing for their interaction," Papadopoulos said. "The thinking around mobile services is at the heart of this transformation and true convergence between the computer-centric view of the Internet and the people-centric view of telephony."

Microsofts Rashid likewise sees the network becoming more aware of individual users rather than their devices. Information about our personal activities, he said, will soon be stored in fewer redundant (and possibly inconsistent) ways, with more of an event- and activity-driven focus on what we need to know.

That evolution, said Rashid, in Redmond, Wash., will continue with growing use of machine intelligence to deduce and meet higher-level needs. When it comes to commuter information, for example, "I can go on the Web and get a nice map, updated constantly with images of intersections," he said. "The problem is that its more work than I want. What I want is something that tells me, You have to leave the office now."

Such personal interactions carry with them a natural concern about managing personal data. Papadopoulos said, in fact, that network identity is becoming the defining issue of online commerce.

"It is the heart; it is the recognition that the Internet has to move beyond naming computers to naming people," Papadopoulos said. "As an industry, we need to formulate the foundations of secure, authenticatable identity openly ... and, most importantly, preserve the rights of the consumers to share or not share them."

IBMs Adkins, in Somers, N.Y., agreed, saying that all such efforts must "make sure that the user has control over privacy. You may not always want the network to detect your location or presence; we want the user to control the ability to be discovered."