The State of Cyberspace

January 1st, 1996

Stroll any street in San Francisco and half the billboards and stickers pasted to its surfaces refer to the Net. San Francisco has emerged as cyberspace's First City, and nowhere is that mantle more clearly evident than at 510 Third Street, a refashioned downtown warehouse which houses, among others, Hotwired, World's Inc., vivid studios and Organic Online - four of the most influential and fastest-growing companies in cyberspace.

The hyperkinetic growth these companies are enjoying isn't limited to the denizens of 510 Third Street. 1995 marked the first year that a sizeable number of corporations attempting to find their feet in cyberspace began to develop a realistic sense of what's required to build a competitive advantage there. With that awareness emerged a keener understanding of the business value of the work of the finest web design and programming houses who, across the board, have seen their billings skyrocket toward the close of the year as their ability to attract the most lucrative and interesting projects accelerates.

Nevertheless, top tier cyberspace design firms remain few and far between. Organic Online, co-founded by Brian Behlendorf who originally led Wired magazine's charge into cyberspace when he laid the foundations for Hotwired, Wired magazine's online presence, enters 1996 as arguably the foremost corporate web design studio worldwide. See accompanying profile. San Francisco-based, which grew out of the Interactive Media Festival's ambitious online effort, is home to some of the most forward-thinking talent in cyberspace and has, as a result, emerged as the VRML (hence, interface design) store to beat, with Munich-based Black Sun carrying the flag in Europe. Dimension-X, headquartered in San Francisco, and creators of Liquid Reality, enters 1996 with a reputation as a Java powerhouse. Reality Factory, the IDG web design skunkworks headed by Chris Tacy, though not as visible as others because most of its work has focused on in-house projects, is home to some of the fiercest corporate applications design talent on the web. The Hotwired cachet continues to attract stars in the making with Carl Steadman, Joey Anuff and Mike Kuniavsky among its impressive recent discoveries. The web efforts of vivid studios in San Francisco, led by Drue Miller, io-360 in New York, led by Gong Szeto, and Obsolete in the United Kingdom have also established impressive reputations as firms able to create virtual identities that get noticed.

These companies' preeminence reflect the abilities of the core design talent who work there - an emerging generation of cyberspace design superstars whose skillsets resist easy classification. Typically in their early 20's through early 30's, these Young Turks demonstrate a rare marriage of technical brilliance and aesthetic artistry. They're in the reality fabrication business, after all. The best corporate spaces and hence the best virtual business models are being fashioned by the most creative designers on the web - artists, in essence - and it seems clear this will always be the case. The fees the best web design studios charge have risen to reflect the scarcity of talent they command. The finest web design shops tend not to consider projects under $100K. And climbing fast. Likewise, the prices the finest web talent can command are spiralling. Proven web talent with a broad range of abilities start at around $65K with packages now on the table climbing to around $150K for the very best, with profit-sharing and stock options to boot. 23-year-olds know all the finest restaurants in San Francisco and Silicon Valley.

Educated consensus has it that Netscape employs the best web programmers on the Net, with Rob McCool particularly, and Eric Bina worthy of mention. The newly found millionaire status of the programming team there has only whetted its appetite to ensure that the stock price is as high as it can be when the heavy, golden handcuffs loosen in four years. Another web powerhouse, Hotwired, is aggressively pursuing its efforts to establish itself as a major, long-term presence in cyberspace. Its recent sale of a 15% stake for $7 million prompted a hiring frenzy that has, in recent weeks, taken its staffing level above the 100 mark. Hotwired now employs more people than Wired magazine.

Meanwhile, efforts by web technology companies to establish a commanding lead continue unabated, with the debate over the VRML 2.0 standard the main issue of the moment. VRML, a 3-D version of hypertext, looks set to take the web by storm when Netscape releases a version of its popular browser in mid-January with a VRML viewer embedded in it. The new spec will provide web developers with the ability not only to design compelling three-dimensional images, but also the means of creating, for example, multi-user 3-D worlds and sophisticated CAD-CAM simulations. It represents a very important step forward in the cyberspace technology arena and there's a perception among the savvier web technology companies that current efforts to settle upon a standard presents perhaps the best opportunity going to "control" a technology and, hence, a viable long-term profit opportunity. But the individual efforts of these companies have been blunted by a group of ten individuals called the Virtual Architecture Group (the VAG) who came together around Siggraph, and whose informal meetings in its aftermath have established its members as the de jure arbiters of the VRML 2.0 standard. They are regarded as beholden to no one.

The VAG typifies the shift on the web away from industry leaders to grassroots influence. Sun, Microsoft, Netscape and HP in particular, resent the influence of what they see as self-appointed guardians of the VRML standards process. But they understand full well the influence of this group and court it assiduously. Microsoft, in particular, has been making strong overtures to individual members of the VAG, as well as to the group itself, in efforts to see its VRML standards proposal, Active VRML, also known as Reactive Behavior Modelling Language (RBML), emerge triumphant over the main challenger, the Moving World's proposal, put forward by Sony, SDSC, SGI and World's Inc.

The RBML proposal is considered strong, but two significant concerns seem set to derail Microsoft's ambitions to control the standard. First, RBML contains some key technical oversights: "Microsoft really hasn't thought about the network stuff," said a VRML designer closely aligned with the VAG. "If you're going to build worlds with multiple users, you simply can't do without it." Second, Microsoft has also made it clear that it wishes to take charge of the VRML spec for a short but undefined period. Concern over both the technical and political ramifications of the Microsoft proposal has the VAG leaning toward a middle course, neither giving in to Microsoft, nor ignoring it, but instead incorporating RBML into the VRML 2.0 spec.

On this front, two possible approaches seem credible. First, treating RBML as a language (as another Java, for example) that sits outside the VRML file format and is an external behavior; or second, modifying the VRML file format itself to give it additional linguistic properties, so that in essence, it becomes more of a hybrid. The Moving World's proposal already adds linguistic features, such as paramaterization - which allows, for example, the ability to define a cube and then instantiate it using color as a parameter - and so adding elements of RBML to the spec is feasible. Technically, the second approach will require much more work on the VAG's part, but would result in a stronger spec since to treat RBML as an external behavior would undercut some important technical objectives, particularly those relating to optimization. A final spec appears three to six months away. Further hiccups may arise as both Netscape and Sun present their VRML 2.0 proposals to the VAG in January.

Whatever the nature of the final spec, there are already whisperings in the financial community that its arrival will present problems for Worlds, Inc, the current 3-D multi-user-world powerhouse, looking to an IPO, whose technology is proprietary. VRML 2.0 will, in essence, allow companies to do for themselves what World's Inc. currently charges handsomely for. Recently, IBM and Sega reportedly looked at World's Inc. technology and decided to wait on the arrival of VRML 2.0. Since the departure of their chief network systems architect three months ago, World's Inc has no one playing an active role in the VRML community.

Further tension on the standards front is reflected in efforts to improve HTML.The recent W3O "style sheet" meeting was attended by every significant player in the HTML arena except Netscape. The foremost web designers consider the advent of style sheets a critical step in efforts to raise the quality of the web experience. The idea of style sheets (common to SGML, the HTML "parent", but not yet to HTML itself) is that many of the design attributes of HTML are far more efficiently handled by style sheets than they are by HTML tags. Currently, HTML describes information structurally, but leaves rendering up the browser, with the result that the wide variety of browsers in use creates a correspondingly wide variety of user views. Style sheets provide a way for web designers to control what users see by establishing consistent styles across all boundaries, whether they refer to point size or coloring, for example. A web site governed by style sheets could contain a "global" site style, with nests of further style sheets specific to individual "pages" or sections.

Netscape's reluctance to pursue the style sheet protocol perhaps reflects the company's stated willingness to see the death of HTML or, rather, its replacement by other data types, while content developers and users still sense HTML is and will continue to provide great strengths. Another disconnect with the user community relates to Netscape's treatment of content negotiation which, if addressed properly, will finally provide a seamless method for users to be served the data types their browsers can "understand." Currently, if a user gets to a page that contains a data type her browser cannot understand, the result's a mess. Few other web technology companies have addressed content negotiation, with no one focusing on it in particular. There is a belief among key members of the web community that Netscape should use its influence to set a standard for others to follow.

Meanwhile, end users press ahead with efforts to push the style sheet standard, independently of Netscape. Netscape would do well to beware grassroots efforts. A recent end user initiative resulted in the release of Apache, a public domain HTTP server for UNIX that is widely heralded as the finest mainstream web server available, indeed the server of choice for web-savvy organizations such as Organic and Hotwired. It is fast, efficient, and very stable. A core strength is the ease with which it allows modules to be added that provide additional functionality. SSL modules (and with them, the ability to secure transactions) are available. Within a matter of weeks Apache looks set to become the leading web server of choice worldwide. For general statistics on web server usage, or to check on the server choice of any url you specify, check out:

Netscape, however, continues to establish a commanding lead among online providers seeking a secure, stable solution backed by the service and support that only a corporate vendor can provide. Netscape Commerce and Netcape Communications servers, in particular, have made significant inroads in recent months with their market share increasing 50% in November alone. Market trends such as these are helping Netscape make sense of its role in cyberspace, as a server technology provider.

Netscape's early efforts to clarify its role and technology in the face of a landscape that changes by the month are reflected in its and other companies attempts to tidy house after the frenetic efforts to establish a strategic foothold in 1995. Currently, for example, Netscape's advanced servers sit atop an Oracle database. Details of the licensing arrangement aren't public, but it appears every "publishing" and iStore server Netscape ships puts a very large smile on Larry Ellison's face. Agreement also, reportedly, requires Netscape to furnish Oracle with information about who is purchasing what. Netscape's current dance with Informix reflects a concerted attempt to find a workaround.

Sun picks up 1995's award for the company who has done most, less wittingly, to make a success of the web. Sun's toolkits made it an easy, early choice for Netscape as a development platform and hence, the server platform it develops for, first. The result? Netscape is singlehanded responsible for not only driving sales of Sun boxes through the roof but, arguably, reversing a sales trend that was, pre-web, lackluster at best. Sun may be more fortunate with Java, however, now that it realizes what it has on its hands. The development language looks set to take the web by storm. Grassroots efforts seem likely to ensure the VRML-Java combination emerges as a durable standard. The best web designers are moving ahead swiftly with efforts to incorporate both technologies into their armouries. We are already, for example, beginning to see Java-based threaded newsgroups. And, unnoticed at Sun, Freeman Murray is writing a distributed, virtual world in Java. His effort is important, because it gets to the heart of what Java and virtual worlds are all about - the ability to create fully distributed, fully composable and, as a result, fully scalable environments in a way that very few Internet technologies allow.

The Big Names are taking Java very seriously, too. Netscape and Microsoft's decision to license the product reflects a belief that the Java approach has profound strengths. Against that, Sun's efforts to ensure Java doesn't get away from it - licensees have to hand over source code for any Java extensions they create so that improvements, whoever makes them, get included in Sun's version - would alone seem likely to spur Microsoft, in particular, to develop or reverse engineer an alternative. Microsoft has all the tools, resources and incentive to take Sun on at its own game, particularly given its proven ability to compete head to head in the marketplace. The battle to own the defining cyberspace development tool is not one Microsoft will ever relinquish.

The Java-VRML story aside, there remains room for a wealth of cyberspace-related tools and applications. Perhaps the biggest area of opportunity relates to efforts to deal with information overload. If you thought you were swamped with information now, just wait six months. It's clear that delivering information is an easy game, and that some of the biggest fortunes to be made in the information technology arena over the coming years will not be made delivering information, but stopping it. A notable early success in efforts to control and self-select a flow of information is led by Architext, whose eponymously titled product takes vector-based retrieval technology and brings it up to date, 90's style. Architext is a retrieval, rather than a database product. It provides an API allowing searches of any type of text-based database content and allows users to navigate their way to information without knowing anything about databases, by typing, for example, "What in the heck has happened to Microsoft Network content providers since August 25th?" Searching is concept-based. Architext is among the first generation of cyberspace-related products that holds out the real promise of "data for the rest of us."

In a sense, the emergence of Architext, and the unique position in which it finds itself, is emblematic of the fact that cyberspace, despite all the focus and activity in and around it, remains a medium we are only just beginning to come to terms with. 1996, then, may well herald the arrival of a larger understanding not so much of the technologies that make doing business in cyberspace possible, but a better sense of how to apply them.

Currently, everyone is floundering. If your company isn't yet making money on the web and cannot readily see a way to make it pay, don't worry, neither does anyone else. And if you think you are clear about how it will pay, best think again. Worlds Inc's need to reassess its business model, the triumph of Apache, and the inability of the VRML corporates to exert control, are small examples of a fundamental shift in the nature of marketplace and, as a result, of the economy itself. The web, quite simply, forces a new business reality upon us all. Which means that if you're basing your business model on the rationalities of an off-web economy, you're in trouble. A general rule of thumb is that what you have been doing so far will need to change. Not, of course, that it's yet clear what to do instead - as Jay Ogilvy, a principal at the Global Business Network, and one of the savviest minds thinking on the subject, has observed: "There's a Nobel prize going to the person who figures out the economics of information." Jay's latest book, Living Without A Goal, ISBN 0-385-41799-3, is worth a close read.

But there are things you can do to move yourselves toward profitability. Invest in people and let them be creative. Allow for the fact that creating a competitive advantage in cyberspace requires a most unusual talent - one which which requires far, far more than a sophisticated command of information technologies. Building a virtual presence that engages clients and prospects means literally manufacturing a workable reality. Suffice to say, the interdisciplinary mishmash of skills required to "make reality work" is about as broad as it gets. Which is why the artistry of the very best cyberspace design teams will reflect, among other things, a keen understanding of design, psychology, information and media theory, architecture, the economics of information, the practice of name just a few.

Needless to say, companies that don't "get" the web, despite the financial resources they might be committing to it, don't yet understand the business of cyberspace; whereas companies who are cottoning on, have a growing understanding that corporate cyberdesign isn't about making something that's an approximation of, or an adjunct to, or something subsidiary to their existing business processes; that a web site is, elementally, nothing less than a virtual business model; and that the ability to create workable virtual business models is emerging as, arguably, the future's most highly prized business skill. They're the companies who are investing in creative people and giving them room to think. You might consider doing it, too. Sooner, rather than later, a competitor will.