No matter who wins the battle of the browsers, the rest of us are likely to lose
It is one of those perfect New England days. Late spring. Sunny but crisp. There is a buzz on the MIT campus. What seems to be the entire student body swarms west, making its way toward the hivelike dome known as Kresge Auditorium. Remarkably, the traffic in the campus thruway known as the Infinite Corridor is going in just one direction, and the usual hallway exchanges about theoretical physics and aerodynamics have given way to gleefully juvenile cries. "He's coming! He's coming!" "Did you see him?" "He's really here!"
A good 15 minutes before show time, empowered-looking students, today acting as bouncers, hold up their arms and turn new arrivals away. The 1,156-seat auditorium is filled to capacity. Press credentials have to be shown several times to gain entry to a small cordoned-off area where more than 100 journalists are scribbling on pads or testing tape recorders.
Beaming, Michael Dertouzos, director of MIT's Laboratory for Computer Science, refers in his introductory remarks to the main attraction simply as Bill, and no one is confused. "Gates" has become as superfluous to the chairman of Microsoft as "Sumner" has to Sting. Dertouzos also reminds the packed and highly charged auditorium that the last time a speaker for the Distinguished Lecture Series packed a hall was when Jim Clark, the chairman and cofounder of Netscape Communications Corp., came to campus.
Dertouzos isn't in Kresge a few hours later when a sparse and almost anemically polite audience gathers to hear another lecturer, Tim Berners-Lee. Like Gates and Clark, Berners-Lee is here to talk about the phenomenal growth and seemingly limitless future of the World Wide Web. The diminished audience is ironic. While Gates and Clark head companies that make products for the Web, Berners-Lee invented the Web and now leads an organization that comes closer than any other to officially controlling it.
But the small audience is also a reflection of influence. For proof, visit almost any home page on the Web and notice how often you come across a gleaming capital N, the name Netscape, or the phrase "Best when viewed with Netscape Navigator." Check out a few more sites, and you'll probably hit on a reference to the rapidly up and coming Microsoft Internet Explorer. But unless you know exactly where to search, you're not likely to run across a mention of Berners-Lee's World Wide Web Consortium--or any of its technical achievements--even though the consortium, if successful, would make Web-page browser icons like the shiny N irrelevant.
W3C, as the consortium is often called, exists to design and promote the adoption of standards for the way information is displayed, modified, and interacted with on the Web--and those who are members join with that implicit understanding. But in reality, the consortium does little of substance. Instead, Netscape almost single-handedly sets the Web's evolving standards--though now Microsoft is making a fast-break bid to wrest that control from Netscape. The result? The Web has thrived, but the consumers and businesses that are coming to depend on it face an increasing number of potential roadblocks caused by growing incompatibilities between the various software programs used to create and access Web pages. And the thousands of companies that make up the Web software industry risk being hampered by the dominance of a single company acting--as do most companies--purely out of self-interest.
How difficult is it to deal with Netscape's de facto standard setting? Anyone who's surfed the Web with a non-Netscape browser--or even with an older version of Netscape's Navigator--is familiar with empty or distorted graphics frames, nonfunctioning interactive features, and a host of other defects that manifest themselves from site to site. Now, Netscape Navigator users are starting to encounter similar problems on the relatively small but rapidly growing number of sites that are optimized for Microsoft's Explorer.
Keeping up with the latest Netscape variations mystifies even seasoned Web masters like Andrew King, principal of Athenia Associates, in Ann Arbor, Mich., who runs a Web site (http://www.webreference.com) devoted to following Web technology, including tracking the ever-changing features of various browsers--one of at least a dozen such sites that have sprung up in the recent past. "We do this all day, and we still have trouble keeping up," confesses King. Ditto for Matt Stevens, a network manager for Wired magazine's on-line spin-off, HotWired. Stevens says his job is to "keep track of the latest and greatest" Web features, but with Netscape it isn't always easy. "You have to dig and dig and dig," he says, "because Netscape isn't really good about informing people about the new features in its latest version. And it's always changing." He routinely checks Netscape's Web site to compare release notes of various versions.
The confusion can be particularly harmful and costly to companies trying to put up an attractive Web site that's accessible to all browsers. In July David Siegel, president of Studio Verso, in San Francisco (whose Web site, http://www.highfive.com, serves as a barometer of state-of-the-art Web design), began circulating within the Web industry an open E-mail missive about the dangers of Netscape's strategy of constantly tossing out new self-proclaimed standards. "Already I spend thousands of dollars every month (real money) making sure our sites look good on Netscape Navigator for the various platforms," Siegel's E-mail said. "I could write a book on the differences among versions and platforms for this one product. I [want] my clients [to] be assured that their money goes into design, not into working around cross-platform, cross-version compatibility issues. I'm happy to build extras for one browser or another, but when the basic fabric of a site is torn by differences in the way browsers approach basic elements of typography and layout, I feel the consortium is not serving end users. . . . We all know this isn't a game, but I fear that in trying to shake off Microsoft, Netscape will end up hurting people like me and my clients in a flurry of exciting new features and overlapping releases."
The W3C is located at MIT, in seven rooms on the third floor of the Laboratory for Computer Science, with a paid staff of approximately a dozen MIT researchers and an annual budget of $3.5 million. (There are also satellite branches at the National Institute for Research in Computer Science, in Rocquencourt, France, and at Keio University, in Tokyo.) The consortium was founded in October 1994, at which point Dertouzos wooed as director Berners-Lee, the shy, self-effacing Englishman who wrote the URL (uniform resource locator), HTTP (hypertext transfer protocol), and HTML (hypertext markup language) protocols that have become the lingua franca of the Web. At the time Berners-Lee was a researcher at the European Laboratory for Particle Physics (CERN), in Geneva. He had grown tired of being bombarded by business executives begging him for specs they could use to shape the Web. He decided to protect the Web's future by moving his efforts to a vendor-neutral environment. The newly formed consortium had little trouble attracting members; the list now reads like a high-tech Who's Who: Netscape, Microsoft, America Online, IBM, Digital Equipment Corp., Sony, AT&T. Most of the 140 corporate members pay annual dues of up to $50,000 for the privilege of belonging to the consortium; smaller companies and nonprofits pay as little as $5,000.
Berners-Lee talks a bold game when it comes to keeping the Web free of single-vendor influence: "Corporate information-technology strategists should think very carefully about committing themselves to using features that will bind them to any one company," he says. "The Web has exploded because it is open. It has developed so rapidly because the creative forces of thousands of companies are building on the same platform. Binding oneself to one company means limiting one's future to the innovations that that company can provide." Consortium public-relations backgrounder sheets are filled with similarly determined statements: "W3C was founded to develop common standards for the evolution of the World Wide Web," reads one, "[and to] provide a reference-code implementation to embody and promote standards."
But the W3C is like a new college graduate with a wealthy and influential family, for whom expectations are high and contacts plentiful, but whose career doesn't seem to get off the ground. The consortium has failed to take control of setting standards for the Web. W3C personnel tend to downplay the failure, claiming that setting standards isn't the consortium's whole mission. "We're an industry consortium that works with its member companies on areas they've approved, one of which happens to be developing standards," says MIT/W3C research scientist Jim Miller. But if the W3C isn't primarily an industry standardssetting organization, it's not because it has other things to do; it's because it's unable to set those standards.
Dertouzos indirectly acknowledges the consortium's lack of results by insisting that the process of trying to set standards is valuable in itself. The W3C offers its members "the opportunity to scream around a table with their competitors," he says. "You know, a great deal gets done that way." However, when you have an explosively growing business to run, in an industry that measures time in Web years--2.6 months per year, according to Berners-Lee--who has time to sit around a table, let alone with rivals?
Certainly not Netscape. You'd never know it, though, from Marc Andreessen, Netscape's cofounder and senior vice-president of technology. "In the networking world, standards are a lot more important than they were in the PC world," he says. "The only way to meet all the requirements for sharing information and for communicating across a huge diversity of systems is through open standards."
And in fact, Netscape has been devoting an increasing amount of resources to the issue of standards. Since its inception, the company has joined at least five computer-related standards bodies and is considering joining four more. Of Netscape's approximately 1,500 employees, director of technology Martin Haeberli estimates that at least 5% are involved with standards bodies--attending meetings, workshops, and conferences, or participating in E-mail discussion groups. Along with the time commitment to those bodies (typical standards discussion groups can generate some 50 E-mail messages a day), Netscape makes a financial commitment. The company won't disclose figures, but in addition to the cost of dues, travel, and development workshops, Netscape recently hired Carl Cargill as its standards strategist, to work full-time on standards issues.
But in the end, Netscape never offers to adhere to standards recommended by the W3C. When pressed, Andreessen concedes as much. "We've been sort of lucky in that most of what we've innovated either is or is becoming a standard," he says. But is it luck? "Well, now that you ask," he chuckles, "what helps us most is that we were first, and we continue to be first--because that's what we try to do."
Since the company jumped into the scene in April 1994, bringing to the Web an exceptionally easy-to-use, full-featured browser, Netscape's business model has been based on leapfrogging the status quo and coming out with new features ahead of anybody else. Standard Web-page features like the ability to center, align, frame, and move text around in chunks were developed and popularized by Netscape. Those innovations spawned an entire new industry of browser developers, Web masters, and value-added Web service providers, and brought the Web into the homes and offices of an ever-increasing portion of the world's population. Of course, groups other than Netscape have contributed and continue to contribute new Web features and extensions that are becoming either standards or for the time being de facto standards, most notably Sun Microsystem's Java programming language and Microsoft's object-oriented development tools known as ActiveX. (In October, Microsoft agreed to turn control of ActiveX over to the Open Group in Cambridge, Mass., a software-industry group that has helped shape standards for the UNIX operating system.)
But few would disagree that the majority of the HTML features and extensions that the W3C ends up blessing--and more important, that most people end up using--come from Netscape. As market-research and consulting firm META Group Inc.'s program director Stan Lepeak points out, "It's Netscape's mark that defines what is and is not a standard."
Netscape managers suggest that the problem isn't that the company is determined to maintain control of the Web but rather that the consortium approach to setting standards is somehow lacking. "You can have more fun whacking yourself with a ball-peen hammer than going to standards meetings," says Netscape's Cargill. Haeberli goes further: "Being a consortium, the W3C is composed of its members," he says. "And so what that means is that the consortium, without its members, isn't. I don't want to dismiss it, but it really doesn't have an existence."
On the other hand, who can blame Netscape? After all, setting standards for the Web has been the company's main competitive edge. If Web users go with another company's browser, they risk using something that simply doesn't work with many sites because it doesn't comply with Netscape's standards, which are set on the fly from Navigator release to Navigator release. The same holds true of the software that runs Web sites on servers. As Vint Cerf, senior vice-president of data architecture at MCI (best known as the Father of the Internet) observes, "There's an understandable tension between agreeing on standards that everyone supports and product differentiation that distinguishes one product from another. Netscape has been pretty focused on product differentiation in its short history, and from the business perspective, that's understandable."
The policy of ignoring the W3C and setting its own standards certainly hasn't hurt Netscape, whose Mountain View, Calif., parking lot is filled with BMWs, Porsches, and Mercedes so new that many don't have license plates. Netscape fetched a $2.2-billion market value on the opening day of its initial public offering while it was still losing money and showed little promised of making any in the near future. Why should it change its business model?
One might assume that Netscape's behavior is a source of considerable frustration to the W3C. In fact, consortium personnel are quite tight-lipped about it. Berners-Lee seems to insist that being a standards hog is incompatible with success. "Companies know that it is only interesting to compete over one feature until everyone can do it," he says. "After that, the feature becomes part of the base, and everyone wants to do it in one standard way. The smart companies are competing on the implementations--aspects like functionality, speed, ease of use, and support--that differentiate products." But if Netscape's and Microsoft's Web software isn't competing on nonstandard features, that's news to them and to everyone else in the industry. Only MIT's Jim Miller (who works in the W3C's "technology and society" area) risks breaching decorum, if mildly, by being forthright: "There are, we have to be honest, companies that while supporting a standard, have a proprietary way of doing it."
Industry players outside Netscape and the W3C are, predictably, more blunt about things. "There's nothing wrong with a company's advancing proprietary standards so that they become de facto standards," says Prakash Ambegaonkar, CEO of Frontier Technologies Corp., an Internet- and intranet-applications software company in Mequon, Wis. "The problem is if it isn't sincere about being open. Duplicity. There is a feeling that Netscape has that." Netscape's ability to come out with new features before they are standardized gives Netscape "a special advantage," he adds. David Gifford, vice-chairman of electronic-commerce software purveyor Open Market, in Cambridge, Mass., agrees. "As much as they say they want to standardize, they really don't," he says.
Netscape doesn't always bother to hide its disdain for industry standardssetting bodies. "Netscape's attendance at open standards meetings has been spotty," notes Eric Sink, who works for Netscape competitor Spyglass Inc., in Naperville, Ill., and chairs a working group within the Internet Engineering Task Force (IETF), an informal grassroots Internet standards body. "While it's gotten better lately, for a long time Netscape gave the impression that it just didn't care." The crowning insult, says Sink, was when Netscape blew off an IETF meeting two years ago in San Jose--some five miles from Netscape headquarters. Even when Netscape attends meetings, he adds, its representatives don't seem interested in advancing the agenda. "It's as though someone from on high said, 'Let's not make a big deal about it, but let's not let anything go by without our knowing about it either,' " he claims.
Web-page designer Siegel becomes livid on the subject of Netscape's track record in standards efforts. Siegel is a member of the W3C and an invited member of its HTML review board E-mail discussion group. He alleges that the group is often "the last to know" about new Netscape features. "It does everything by press release," he gripes, "and conveys the message that 'we're not going to play by the rules.' " He claims that when he recently ran a page on fonts for the W3C, Netscape didn't contribute and that the company has ignored standards-setting work on style sheets. Netscape was conspicuously absent from a Web conference in Versailles last spring, he notes. "Netscape said it was too busy, which is its reason for everything," grumbles Siegel. "If Netscape is not acting as part of the consortium at this low level, perhaps it's time the W3C asked it to either be in or out of the formal standards process."
Indeed, why doesn't the industry band together to censure Netscape in some way? The answer is probably obvious and is vividly answered in an E-mail from Fred Baker, senior software engineer at Cisco Systems Inc. and current chairman of the IETF. "Meddle not in the affairs of dragons, for you are crunchy and taste good with ketchup," he wrote.
But it's not just fear of being crushed by the Netscape steamroller that keeps companies and individuals from protesting more vociferously. The fact is many, perhaps even most, people are quite happy with Netscape's de facto standards setting to date because the company arguably has set good standards. "Consider the alternative," says META Group's Lepeak. "Were it not for Netscape, we'd all be stuck with last year's browser."
Most people right now are simply concerned with getting functionality as fast as possible to the Web, he contends. He dismisses those who are upset by Netscape's actions as either Internet "purists" or Microsoft "loyalists." Many Web enthusiasts are rabid Netscape devotees, like Pepperdine University's Internet researcher Ogden Forbes, who wrote his doctoral dissertation, at the University of San Francisco, on the history of the Internet. "Netscape has given a major gift to legions of people worldwide," he gushes. "Thank you, Netscape!"
And it's true, at least in the short term, that having a single company take control of an emerging technology can provide much-needed direction and leadership. But the longer-term ramifications of unilateral standards setting is more grim, to judge by history--and one need hardly look outside the computer industry for evidence. Has anyone noticed companies rushing to restore IBM's several-decades-long dominance of the corporate computing world, maintained largely by its having set technical standards for computer hardware? Does a large percentage of today's PC users appreciate the decade it took operating-systems standards-setter Microsoft to offer DOS users a workable graphical user interface (Windows 3.1) and the additional half decade to bring out a successor (Windows 95) that equaled the functionality and usability of the Macintosh operating system of the mid-1980s?
The essential question is this: Once a company firmly controls standards, what's to keep it working hard to make sure that new features benefit customers rather than head off competitors? The question becomes all the more critical as browsers evolve from operating system add-ons to the actual core of the operating system--a direction that both Netscape and Microsoft have made no secret of taking.
But the Web industry and those of us who make use of its products don't seem to be in a position to do much right now besides cast a vote for either Netscape or Microsoft. Perhaps the best we can reasonably hope for is that the two will end up sharing control, each forcing the other to keep its offerings in its customers' best interests. What doesn't seem likely is that either will give up the battle and throw its weight behind jointly developed standards. As Clyde Seigle, vice-president of technology development for tax and legal information provider CCH Inc., in Riverwoods, Ill., notes, "The Internet has come a long way from the days when everyone was supposed to cooperate."
Alessandra Bianchi is a contributing writer for Inc. magazine.