CNET también está disponible en español.

Ir a español

Don't show this again

HolidayBuyer's Guide
Internet

Universities face challenge

Universities have long complained that online congestion was stunting research, and the government may finally help them. But some question the wisdom of spending millions of taxpayer dollars to build a new Internet exclusively for academia.

Jack Hodges isn't upset that he has to share the Internet with millions of other people. It's the way they use it that bothers him.

"My complaint is that the vast majority of users are Internet illiterate or inarticulate," said the professor of computer science at San Francisco State University. "It is a person's inability to use a tool properly that makes that person an impediment to others. Students and many non-professionals--and professionals--dump every little idea and whim onto the network."

The result: "Bandwidth goes up, performance goes down."

His concern is hardly new. Ever since academia lost its private domain to the commercialization of the Internet, professors and scientists have complained that they were losing vital research time to unprecedented congestion. Now, however, university researchers believe that they are on the verge of getting the support they need to once again have a worldwide network all to themselves.

But just as their plight is drawing government attention, some are beginning to question whether hundreds of millions of taxpayer dollars should be spent to build a new Internet exclusively for universities, particularly at a time of shrinking budgets and growing deficits. Experts outside the academic community say private companies, not universities, are better equipped to develop technologies that will solve the Internet's problems and allow it to expand.

"We are no longer in a world where we need the government to play start-up for internetworking and computer communications," said Mike Godwin, staff counsel for the Electronic Frontier Foundation, a nonprofit that works to protect privacy and free expression online. "The notion of dealing with [technology development] as a research project, I think, is misconceived."

"I bet if the government offered the same amount of money to Internet service providers, they would get it done faster and better."

Such criticism could not come at a worse time for the universities. Just last week, President Clinton announced a $600 million program dubbed "Next Generation Internet" that will upgrade university networks and contribute toward Internet II, a proposal that would give researchers exclusive access to high-speed networks, allowing them to share data and test new networking technologies without getting stuck in online traffic.

Without this network, researchers maintain, companies will make significant improvements to the Internet as we know it but not lead the kind of technical revolution needed to take us into the next generation for which the White House's plan is named. "Researchers are not resentful of the commercialization of the Net," said David Hitlan, spokesman of the Babar project, a high-energy physics collaboration involving the Stanford Linear Accelerator Center at Stanford University and 75 other institutions worldwide. "It is just a shame that such a useful tool cannot be exploited by scientists."

Hitlan says the Web used to be a perfect medium for discussion groups, memos, and videoconferencing. But the performance of the network has steadily deteriorated since the Babar project's formation in 1993.

The project tries to work around these glitches, but solutions are frustratingly elusive. "We have tried to get around [the problem] by having mirror sites. This is difficult because our research requires password entry, which forces the researchers to log on from a specific computer," Hitlan said. "We have particular difficulty connecting with colleagues in the United Kingdom, but Europe is a problem in general."

The problem came to a head last March. Researchers from the SLAC program sent a memo to the Federal Networking Council in Washington, decrying "catatonic" connections that made "it essentially impossible to log in and control execution of jobs across the network."

What they want, according to Mark Luker, program director for the National Science Foundation's NFSNet, is a network to not only allow communication but one to "research and develop new networking applications, to determine what kind of a structure we all need."

The new government money--if it is approved--would be invested in projects such as I-WAY, an experimental high-performance network linking dozens of the country's fastest computers and advanced visualization environments. One projects funded last year, for example, examined wireless local area networks that would allow conference participants and offices to communicate with each other without being physically hooked together.

But critics say that, even if universities need to upgrade their networks for experimental research and to maintain community, the government should not be doing the work.

"You can take the same money the government would spend to build infrastructure and put out a bidding process and see who can meet your needs," said Godwin of the Electronic Frontier Foundation. Even some within the academic world agree. "The proper way to create new infrastructure is to let the commercial sector do it," said David Farber, a professor of telecommunications at the University of Pennsylvania. "The kind of high-speed networks the research community needs are commercially available."

The problem with Clinton's proposal, Farber said, is that "it doesn't give us a next-generation Internet. It simply upgrades the university networks to highest-speed networks commercially available."

What researchers can do, according to Farber, is find technology that goes above and beyond what private companies are working on. "There are a lot of people in the private sector who can do good networking technology and who have very good ideas about applications," he said. "I am not quite sure that the university is in the driver's seat for stuff which is commercially available."

It is here that many see the university's most valuable role: doing independent research not gauged by the bottom line of any Silicon Valley accountant's books.

Raman Khanna, the director of distributed computing and communications systems at Stanford, says academic researchers should not just find ways to relieve congestion on the Net. Universities should look at currently "unimaginable" internetworking, he says, because most companies aren't willing to take the chance on the type of hare-brained ideas that sometimes yield scientific miracles.

"This is an ongoing cycle," Khanna said. "The community feels that the technology is not mature enough. Researchers and universities as a whole are more willing to experiment, try new things.

"As they work out, we can transfer the technology to the commercial Internet so that others can benefit. Internet II is not the end. There may be an Internet III."

 related news stories
ISPs benefit from college cutbacks
Colleges charge students for Net use
ISPs fight for their lives
Banding together for bandwidth
Grants to speed research
Technology issues go to school
Is Clinton's Net plan for real?
Clinton to expand Net capabilities