My SXSW Swag: A State of the Web Address

O'Reilly's Web DevCenter

Interdependence doesn't mean that everyone can be a lot less dependable, but rather that we all represent the potential of being the weakest link in the chain.

Even though I've been back from South by Southwest for days now, a few themes from the conference still haunt me.

SXSW is a legendary gathering of some of the finest minds in independent film, music, and the "interactive" industry, which in recent years has come to mean "Webfolk." Hosted by the lovely town of Austin, TX, where live music, great food, and liberal politics walk hand in hand under the watchful eyes of the Capitol and the UT Clock Tower, SXSW stands alone as a sort of strange testament to interdependence and harmony between disciplines.

Where other conferences focus on providing strategy, or techniques, or platforms for vendors to ply their wares to a single industry, SXSW invites the masters of the visual, the musical, and the interactive worlds to think hard during the day, then continue to think hard but party harder during the warm spring evenings.

When you can buy a share of iXL or Razorfish stock for less than you'll pay for a 12-ounce can of Coca-Cola in a hotel at a conference, it may be hard to imagine that the denizens of the Web industry would be upbeat. But oddly, that was the case. Certainly, there were signs of the slowdown, concerns that the industry that made it possible for tens of thousands of us to shed the skin of unemployed or underemployed liberal arts graduates or corporate hallrats was going through a radical transition phase, bringing a future that looks cold and forbidding to CFOs and market analysts everywhere.

But for the most part, SXSW is composed -- speakers and audience both -- of thoughtful and forward-thinking idea machines. This year, however, saw as much earnest reflection on how to avoid the mistakes of the past as how to exploit the potential of the future.

The Internet culture was born out of the seemingly bottomless pit of defense spending of the fifties and sixties, and formed in environments where, more often than not, someone else was footing the exorbitant bill for high-speed 24/7 access to alt.devilbunnies. As such, it has a tendency to believe itself capable of reconciling the often conflicting goals of free-form expression and creative ferment on the one hand, and meeting the bottom line on the other.

My choice of reading for the plane ride out from Raleigh ("Dealers of Lightning," Michael Hiltzik's compelling tale of Xerox PARC) seemed at once a perfect parallel and counterpoint to the exuberance of the Web generation soon to assemble and an archaic document from a far simpler time. I'd argue that the coming decade shares much with Bob Metcalfe's "best efforts" networking system, Ethernet, developed at PARC, which Alan Kay called "one of the great finesses of all time, an object lesson in how to make something work when you don't know how to make it work well."

I'm not sure why, but that phrase resonated through my head for the entire conference: "how to make something work when you don't know how to make it work well." It's as though our industry needs just such an object lesson, a reminder that sometimes you have to make do with what you have and see if you can succeed in meeting your objectives, above the clamorous din of objections on the grounds of inelegance and wastefulness or flying in the face of established traditions.

Sometimes you just have to make it work, even when you don't have the faintest idea how to make it work well, and everyone around you who tries ends up as a barely-remembered branding exercise above a "for rent" sign in SoMa or Cambridge or Silicon Alley or Austin's new media strip.

Ripples travel faster in a connected medium.

For those who don't know how Ethernet works, perhaps a metaphorical overview is in order.

At a time when networking generally consisted of big iron connected in a star topology, with dedicated lines stretching out to multiple individual terminals, Metcalfe decided that his networking system -- which needed to let computers share a communications channel to talk to each other (itself an innovation) -- should draw from an experimental network called Aloha.

This older system used radio signals in a scheme that allowed stations to share a frequency. Each station could send whenever it needed, and collisions (where two stations sent messages simultaneously, garbling both messages) were resolved by having each station "back off" and resend after a random time interval had passed. The assumption was that even without knowing when the next transmission would start, any station had a pretty good likelihood of successfully sending a transmission before another began.

Metcalfe improved on this method in a number of ways: by building receivers that "listened" before they tried to send, to make sure the channel was clear; by adding a better backoff algorithm to the mix; and by formalizing and publishing the specifications for the actual physical layer (the wire) through standards organizations. In essence, Ethernet's triumph lay in exactly the things that made it so objectionable: It succeeded through allowing a certain amount of noise and failure while making it possible to economically transmit more messages successfully over a shared medium whose mechanics were known to all. By allowing for, even anticipating, a certain amount of failure, it made it possible to succeed at something new: the exploitation of a shared communications medium.

Why am I obsessing on Ethernet, of all things, when you're probably waiting for the scoop on the SXSW Web Awards, or whether there were as many great parties as last year? Put simply, in the model for Ethernet lies the secret to understanding how the Web industry is going to survive despite the shakeouts of the past six months. When an industry such as ours, fed by dreams of incredible wealth and a revolution in the ways that businesses communicate with themselves, with each other, and with their customers, suffers from an inability to achieve perfection, perhaps it's time to consider simply making things work when we can't make them work well. Besides, I arrived in town late and missed the mother of all SXSW parties, at Lane Becker's house. So, no pictures or stories to speak of.

Earlier, I spoke of what I saw as the refreshing focus on the future that characterizes SXSW as a conference, as opposed to the focus that other conferences have on the present or on communicating the achievements of the recent past to legions of folks desperate to catch up. In my mind, this was the most indicative -- and important -- thing about this year's SXSW.

The attendees and panelists and speakers were there to talk about the potential that still exists in the Web as a platform and as an environment. I saw fear and doubt and concern on the faces of my friends and compatriots, but more often I saw a process of realistic and cold-blooded and yet hopeful assessment of what the future still holds, a willingness to believe against all odds and in the face of dire predictions in the press that the Web still holds much to be explored, much to be exploited. And even if the insane market valuations of the past four years are gone forever, there is still hope that many will be able to make a living wage by improving upon the things that worked in the past, or showing how the tactics that failed can be replaced by those that work.

It's important for us to realize that when the media talks about "dotcoms" as though they are a plague soon to be eradicated through proper hygiene or a band of wild mustangs soon to be broken and yoked to the plow, they don't necessarily mean "Webfolk," who are a different breed altogether. When they talk about the spectacular failure of the new Web consultancies, they obscure the important message: that much of the failure of the Web consultancies is due to the fact that the past two or three years has seen an unprecedented adoption, at every level of the enterprise, of Web technologies, often maintained or implemented by in-house teams or by traditional IT consultancies forced to adapt to the new medium or die.

The Web has succeeded beyond the wildest dreams of even the most rabid of WIRED reading, VC courting entrepreneurs. Even those who did not recognize early that "euphoria is not a business strategy" (to quote Louis Rossetto) have come to terms with the fact that their confidence in the Internet as a world-changing medium wasn't foolhardy, even if their vacuous business plan may have been.

The weeks leading up to SXSW saw the formation of several new companies, such as Adaptive Path, founded on a belief that the Web can be better, can still evolve to meet the needs of its users, even as some of the groundbreaking companies in the same space announced that they were closing their doors, companies like Argus and Associates, whose employees and founders did perhaps more than any other company to legitimize and systematize the practice of information architecture.

Even as the big Web consultancies like iXL and MarchFIRST and Scient and Sapient announced layoffs, people were still excited about the Web as a medium, saw the layoffs as a welcome correction to the madness that the dotcom market drove with a fury that knew little to no reason. As co-owner of a company that focuses exclusively on the Web and Web-related technologies, the correction made it possible for me to hope that inflated and unrealistic salaries would return to normal, that the labor pool would broaden and mature. We still get resumes from people who do not seem to recognize that claiming the value of their stock options in now-defunct companies as a necessary part of their compensation is foolish at best, an indication of an unhinged mind at worst, but the worst appears to be on the wane.

Web developers and designers are paying more attention to Web standards like the DOM, HTML4, XHTML, and CSS than ever before. Many are redesigning their personal sites to abandon their reliance on table-based markup, in attempts to finally learn the power of DIVs and SPANs and positional styles.

The Web Standards Project's recent Browser Upgrade Initiative, though lambasted by many as yet another attempt to encourage design for specific browsers and discourage a universal Web, is gaining recognition with respect to its well-intentioned aim of restoring a platform that allows for universal development, but this time based on W3C standards rather than proprietary implementations in market-leading browsers.

Developers are recognizing the import of the separation of document structure and semantics from style, and many are experimenting further with the separation of structure from presentation, using technologies like XML with tools such as AxKit and Cocoon and server-side XSL to perform transformations. Peer to peer is gaining acceptance, if not as a revolutionary sea change in the fundamental makeup of the network, at least as a powerful new tool for expanding the audience base and exploring new forms of document delivery and distribution.

It's a great time to be alive. And a great time to be on the Web, creating the Web, making the Web work. And that was my swag from SXSW.