Six years after the federal government proposed creating a second internet that would leapfrog the first with speed and technology, most users - business and consumer - are still saddled with the low speeds, transmission delays and the other quirks of today's Net.
This is the age of Internet time, right? So, where's that gigabit-fiber-to-the-PC, priority-packet, streaming-three-dimensional-video Ubernet? And how about the millions of public dollars spent so far on advanced networking research?
"There have not been a lot of tangible results so far," says Scott Rayder, senior technology policy analyst at the conservative Heritage Foundation think tank and a former aide to the House Science Committee, who helped draw up funding rules for the projects. "We have not seen, to this point, public return on this, and a lot of people are concerned."
Give it time, say the folks who have been working on projects such as the university-driven Internet2 and the government-sponsored Next Generation Internet (NGI). You may live in the Internet age, but the advances these two groups are trying to make are the stuff of science and experimentation. Like the wine in the old Orson Welles ads, they won't be ready before their time.
"This is about incremental progress in these technologies," says Sally Howe, associate director at the National Coordinating Office for Information Technology Research and Development, which oversees interagency work on computing matters, including advanced networks.
Although those inside the projects acknowledge there are few - if any - tangible "deliverables" that the public can see as a result of Internet2 and NGI, they say advances are being made on very tough problems and, eventually, results will come. It's just not easy to say what results.
"We have some pretty outstanding efforts now, but what's come of it? It's a little hard to say, because we have so many initiatives it is hard to pick out the ones that are going to be the most important," says Doug Van Houwelin, president and chief executive of the University Corporation for Advanced Internet Development, the university consortium running the Internet2 project. Who knew, for example, that a student project in the early 1990s would create the Web browser, which would result in a boom in the use of the Internet by hundreds of millions of PC owners?
After all, says George Strawn, deputy director at the National Science Foundation (NSF), it took more than three decades for the original Internet to go from idea to what researchers refer to as the "commodity Internet" - the one most of us use. The ARPAnet was conceived as a research tool, and it was "our great fortune" that it had unexpected commercial and educational uses that created unforeseen benefits to society.
"If you raised the same question in 1974 - asked people working on the ARPAnet, 'Is the public investment in the [project] worth it?' - I think they'd have had to dance a bit for you," Strawn says. "And I think we'd have to dance a little bit for you today."
That's not to say that researchers in government, academia and industry aren't making strides in creating a more advanced Internet.
There are currently two distinct high-speed fiber-optic backbones spanning the country as a result of projects begun by the government in the mid-1990s. Hundreds of universities and a handful of corporate partners are tied into those backbones. Separately, a gaggle of high-profile agencies have constructed their own advanced networks, including the Department of Defense, the Department of Energy, the National Aeronautics and Space Administration and the NSF.
While the major backbones are both operated by private companies, there are connection points between them around the country, and certain users on those backbones can also connect through them to the government's advanced networks. While the experimental nature of the networks means they don't have gateways to the commercial Internet, the goal is for the work on all of the networks to lead to technical improvements that can someday migrate to the commercial sphere.
A few corporate partners have already taken technological advances from the work on Internet2 and the NGI projects and placed them into their most advanced commercial backbone offerings.
"The answer to your question is, yes, we are making progress, and the reason it doesn't show in some dramatic way is that it is down in the core of the Net," says Vint Cerf, one of the fathers of the Internet and a member of the President's Information Technology Advisory Committee.
Researchers also caution that the efforts under way to create improved network technologies and applications are not intended to create a sort of "replacement Internet" that will be put into service all at once. Rather than some kind of giant switch being thrown one day, at which point the old Internet dies and a new one replaces it, the process will be far more gradual, as new technologies begin to take over for existing ones.
"It's quite impossible to ever have a 'flag day' again - the Internet is far too big to have one," says Steven Bellovin, an AT&T fellow, referring to a complete switchover on a given day. "The last flag day was Jan. 1, 1983, when we converted from the old ARPAnet protocol to TCP/IP [Transport Control Protocol/Internet Protocol]. You can't even contemplate such a thing now."
Most who are old hands when it comes to the Internet will tell you there's always change and experimentation happening. But the formal idea for working toward some sort of separate, second national network began around 1994, when the NSF prepared to turn over control of the NSFnet - the last public remnant of the original Internet - to commercial management.
Commercialization left some of the Internet's initial clients, namely research universities and government laboratories, without the data playground they had long enjoyed. When the government permitted commercial use of the Internet in 1991, it quickly gobbled up the available bandwidth once relied upon by original Internet users.
Even if the overloaded commercial Internet could handle the size of their demands - which grew less feasible with each million new users - researchers judged it unlikely that consumers and businesses would put up with a network upon which potentially damaging experiments were constantly being run.
"A lot of these experiments you can't do on the live Internet, because you can't risk the interruption of communication to people," Bellovin says. "It's easy to put together something that looks fine on paper, but it doesn't work in the real world."
In fact, researchers sometimes "wall off" a portion of the experimental backbones and deliberately run destructive programs to see what happens and how things behave. Researchers have discovered that the basic Internet protocol, TCP/IP, behaves in unexpected ways when it is used in enormous, ultra-fast backbones.
The real problem was that nobody could really afford to create a new research network alone. Universities didn't have the cash; indeed, many of them have obtained grants from the NSF to pay for their advanced Internet connections. The government couldn't afford the kind of bottomless-pit funding for basic research that marked much of the Cold War era in areas such as atomic weapons or the space program. And private industry is too focused on short-term profitability from its research to build a test bed with a 20-year horizon.
"It's a lot easier for a couple of kids in a garage to start a microcomputer revolution with chips than for those same kids in garages to compete with global network forces," Strawn says. The barriers to entry are simply too great. "It can happen, but not in an Internet timeframe."
So with some federal money as a catalyst, a consortium of universities to do much of the research and corporate partners willing to exchange money or equipment for the chance at firsthand involvement in the research, the U.S. has embarked on its quest for a better Internet.
