SAN DIEGO- Ray Ozzie, creator of Lotus Notes and Groove workgroup software, believes that the Internet is in need of rebuilding. He's not talking about Internet2 or accelerating movement to IPv6, which has its own set of cost and compatibility issues given the installed base of equipment and applications supporting today's IPv4 Internet. Ozzie contends that either via operating systems or middleware a new set of functionality is required to move the Internet to the next plateau within the next five years.
"We can't nuke the Internet infrastructure overnight," Ozzie said, speaking at Mark Anderson's SNS Future In Review Conference here. "We need to take the learning from the last 10 years and rebuild the Internet up one level."
Rebuilding the Internet up one level will require bringing new standard elements into the Internet stack. Ozzie gave examples of what could be part of a rebuilding effort, including XML packets replacing IP packets, encrypted V-cards instead of the usual DNS, a message-switching layer more like the open source Jabber than SMTP (Simple Mail Transfer Protocol), or a dynamic DNS that works more like instant messaging for finding people.
It's far more likely that these kinds of improvements will take root at a middleware rather than operating system level. It takes years for operating systems to design, test and implement new features. In addition, the middleware approach, like Web services, allows components to operate on disparate hardware and software platforms, independent from the infrastructure.
Ozzie also took some shots at e-mail, saying that it's not good for interacting with people, even though it is the dominant collaboration tool. Ozzie has spent his impressive career trying to get users beyond e-mail, so he does have an established bias, but it's hard to argue with his assessment. The signal-to-noise ratio of e-mail clearly favors noise, especially given the spam, corporate e-mail overload, and virus problems.
"E-mail is 30 years old. Some defensive improvements are being made to deal with spam, and there are some marginal improvements in the user interface," Ozzie said. "But, e-mail won't solve the communications problem. We should look more to blogs, instant messaging, and shared workspaces--like Groove--to create a new paradigm for different kinds of interaction."
Another critical part of taking friction out of interacting on the Internet is creating standard authentication and security models that work across the more decentralized architecture of the Internet. Ozzie pointed to firewalls as a major problem that disrupts cross-enterprise communication, and gives a false sense of security. Given that most cyber security breaches come from insiders, that may be the case, but right now firewalls are one of the few lines of defense.
Ozzie believes that the industry should embrace a more "cellular" model for trust and authentication instead of the current hierarchical security model. "Today, each enterprise has its own directory. The different directory products allow for delegation and assume that trust is hierarchical--meaning the boss authorizes employee access levels --but that's not the way people interact across increasingly decentralized enterprises or within peer groups," Ozzie said. Groove has built a more cellular-oriented trust and authentication model to get around the firewall problem in its own domain, but it won't be easy to go beyond that realm.
Another issue that Ozzie brought up was the lack of a standard way to synchronize all the devices that a person uses. "No one likes synchronizing, it's an explicit operation. It should be securely and transparently done whether inside or outside the firewall," Ozzie said. Groove supports synchronizing multiple devices automatically, but this kind of federated capability needs to be a standard that doesn't require Groove or special software for each device.
"The use of technology and devices is bifurcating dramatically, both in the kiosk mode of use where you walk up to a PC and use a browser to get at e-mail-- the 'end user to one device' mode--and at the opposite extreme in the 'n devices (such as phone, PDA, PC) to one person mode,' " Ozzie said. "We want to federate the devices as a logical unit, so they can be auto-synchronized and you can use the devices interchangeably. We need security models to address both models, and in different ways."
After talking with Ozzie, it dawned on me that the conversation was similar to one I had with him a dozen years ago. At the time, Lotus was getting flack, from Microsoft in particular, for developing proprietary software in Notes--meaning Windows lacked the functionality and would rather diss its competitor than praise it for pioneering work. Ozzie and team created system software for its workgroup software because it didn't exist as part of the Windows operating system or any other platform. Over time, many of the concepts and functionality made their way in one form or another into operating systems and middleware.
Ozzie and team have been repeating history over the last five years in developing the Groove collaboration environment, which like Notes is ahead of its time. While Groove was able to leverage more existing code and standards than in the Notes development process, it is showing the way with its synchronization, trust, and authentication models, which could benefit the entire Internet ecosystem.
As Ozzie professes, his goal is not to gain competitive advantage in delivering specific technology, but to integrate that technology--no matter its source--to solve specific customer problems. He's optimistic the components his team has built, or at least the concepts, will migrate to middleware and operating systems. "It can happen, and will happen, it's just a question of when."
Notes took many years to catch on, and even then it was primarily as an e-mail platform. With the Internet as the underlying transport and community environment, it just might be time for Ozzie's vision of pervasive collaboration to take off.
