"Roman Candle" turns 20: Secrets of Elliott Smith's accidental masterpiece (slideshow)
Elliott and the friends with whom he recorded in middle school in Texas (photo courtesy of Dan Pickering)
Richard Lang looks out of his lawyer’s office at the bright orange Coast Guard cutters clipping the San Francisco Bay. It’s a chilly morning in the middle of October, and Lang, the CEO of a video software company called Burst.com, has spent the last hour talking about unfair competition in the software industry and about how he became a victim of Microsoft’s vicious effort to dominate digital entertainment. But he’s been distracted the whole time, pointing out the choppers and the sailboats. You get the sense that Lang is vaguely embarrassed to be here, that he would rather be out there on the bay, or somewhere else far away, working on ways to improve digital technology rather than explaining why he’s suing the biggest company in the industry.
Richard Lang is suing Microsoft because, he says, the company violated Burst.com’s patents covering video delivery on the Internet and engaged in “unfair competition” to prevent Burst from gaining a foothold in the market. Lang says that in the late 1990s, as software companies battled each other for dominance in the emerging world of digital media on the Internet, Microsoft colluded with its main competitor, RealNetworks, in an effort to prevent Burst and other firms from entering the market. And then, after it had sufficiently weakened Burst, Microsoft “stole” its intellectual property, says Lang, incorporating in its latest Windows Media Player the video-streaming technology that Burst had developed — and had shared with Microsoft under nondisclosure agreements — during more than a decade of research.
Microsoft denies any wrongdoing in the case and has vowed to fight Burst’s allegations. Since 1998, Microsoft has been named in dozens of patent-infringement lawsuits, and the company — which has lost almost none of them — has suggested that many of the plaintiffs are failed ventures whose claims are at least partly motivated by the attractive target posed by Microsoft and its $39 billion in cash. Burst.com is one such failed venture. The company is still around, but its prospects are far slimmer than they once were, and the main reason for that, Lang claims, is Microsoft’s bullying.
There’s some debate among insiders in the streaming-media industry over whether Lang has a legitimate, actionable reason to be sore at Microsoft, or whether he is, as a few privately suggest, an opportunistic, litigious hoarder of patents. The courts will have plenty of time to figure that out; his case, which was filed in June, is expected to last — if it’s not dismissed or settled first — at least until 2004.
But Richard Lang’s personal saga — the tale of how he came to see the future of video on the Internet, to form a company around that vision, and to nevertheless fail to make a bundle on his ideas — is just a small part of the larger, ongoing fight over who will control media standards in the digital era. It’s not hyperbole to say that in the future, when all the media we consume is digitized, the company that controls the media formats and the players to decode those formats will have enormous influence over the culture at large — and it’s no secret that Microsoft would like to have such influence. During the past couple of years, Microsoft has poured $500 million into the Windows Media set of technologies, a staggering sum considering that Microsoft has made no money from the software yet. In some ways, the company is pursuing the same strategy it used so successfully during the Web browser wars, marshaling its formidable resources against competitors and flooding the marketplace with good, free technology — and perhaps edging very close to the line of what’s legally permissible.
Microsoft’s misdeeds in the browser fight are well known; they formed the basis of a federal court’s decision, in 2000, that the company was a monopoly and had violated antitrust laws. And now it looks as though that misbehavior will form the basis of Burst.com’s case. On Oct. 24 in Baltimore, U.S. District Judge Frederick Motz, who is overseeing the preliminary motions in Burst’s case, ruled that to prove its case against Microsoft, Burst can use some of the “findings of fact” handed down by the court in the antitrust suit.
Microsoft says that Motz’s ruling, which didn’t indicate how many of the 412 findings Burst can use, doesn’t significantly hurt its chances in the suit. What’s clear, though, is this: All the while Microsoft was fighting, in court and in the media, a charge that it was an unfair monopolist, it was throwing its weight around the world of digital media. Microsoft had to make sure, Lang says, that a small firm like Burst couldn’t stand in the way of Microsoft’s efforts to integrate media delivery into its operating system.
“They didn’t want some little company putting a tax on Windows,” Bruce Wecker, one of Lang’s attorneys, said.
And in the end, Microsoft succeeded. For all practical purposes, Burst has gone bust. Lang maintains that the company is still licensing its technology to set-top box firms in Korea, and that some American hotel chains are providing video through its software. But the company has vastly slashed its payroll and is struggling to stay alive; meanwhile, deep-pocket winners are poised to dominate the future of media. It’s not inconceivable that sometime in the next 10 years, all the media delivered to your computer or your digital TV will come in the form of a Windows Media file. And if that happens, we might look back and see Richard Lang’s failure as the event that laid the foundation of a proprietary media world.
Few people who watch video on the Internet have ever heard of Burst.com. Even folks who work in the “streaming industry” and call themselves experts regard Burst.com as a relative newcomer to the field. RealNetworks, which became the early leader, and Microsoft, which put a lot of resources into catching up with Real, dominate the delivery of audio and video on the Internet. But Lang says that his company has long had admirers — including the rock group U2.
In 1990, U2 became minority owners in Burst.com — then called Instant Video Technologies — with an investment of $2 million. The band wanted “to give something back to [our] fans,” U2′s lead guitarist, the Edge, said to the Hollywood Reporter at the time. But the band was also charmed by Lang’s long view. In 1983, Lang had co-founded, with Terren Dunlap, a company called Go Video, which sold a then controversial (now passé) dual-deck VCR. By the late 1980s, Lang says, he was certain that videotapes were soon going to become obsolete, “and my focus started to shift to what would follow them.” His determination that magnetic tapes would be replaced by electronic pulses of video traveling through networks — even though such networks at the time were more a pipe dream than a certainty — attracted investors, like U2, who liked to think big.
The band members and other investors were told that the company didn’t expect to become an overnight success. “What we said to U2 at the time was to expect no technology for at least five years,” Lang says. He explained to investors that although the company had some grand concepts for video delivery, the specific technologies still had to be invented. Back then, indeed, Lang had no idea that he would be sending video over the Internet — satellite, cable, online services, and the vague but much hyped 500-channel “information superhighway” seemed the best ways to do it.
In 2000, in an effort to signal to customers that the firm had something to do with the Internet, Instant Video Technologies was re-christened Burst.com. (“Talk about bad timing,” Lang says now of the name change.) The firm was named Burst because “bursting,” as opposed to “streaming,” is central to how Lang, in the early 1990s, determined that media should be delivered over a network as inherently temperamental as the Internet. Lang says that streaming data over a network, the standard way of sending Internet video since the mid-1990s, is conceptually identical to television or radio broadcasting: A video is played at the TV station, the pictures are encoded and pushed over the airwaves in real time, and when they get to your TV, they’re decoded and projected onto the screen. In streaming, everything is digitized, and the airwaves are replaced by the Internet.
But Burst.com saw this model as flawed, because the Internet isn’t like the airwaves — it’s more like a system of suburban freeways. The route between any two computers changes from moment to moment, becoming free at one instant, clogged the next. Lang’s technology recognizes this fluidity — and capitalizes on it. When the connection between the broadcasting computer and the receiving one is free, the server “bursts” a large segment of the video to the viewer, which can store the extra information on the computer’s hard drive. The system is smart and elegant, and it leads to significantly better video — so good, in fact, that Microsoft, which started out streaming, now uses a variation of bursting in its latest video player.
But Burst.com’s media-delivery method did not catch on in the marketplace. The company’s researchers spent most of the early ’90s researching the burst technology and, Lang says, “building up our patent portfolio.” That deliberateness seems to have been Burst’s first, and perhaps its pivotal, mistake. Instead of rushing to market with something that worked but wasn’t perfect, it waited.
And Burst was upstaged: In the summer of 1995, a Seattle company called Progressive Networks announced it had solved the problem of delivering audio over the Web. Its software, called RealAudio, compressed sound files and “streamed” them over the Internet. The audio quality wasn’t very good, and the player often skipped and sputtered over a dial-up connection, but the company, founded by a former Microsoft executive named Rob Glaser, was an instant hit. Many major radio stations and networks soon signed on to its vision of streaming media. (In 1997, Progressive Networks, which by this time was delivering both RealAudio and RealVideo, changed its name to RealNetworks.)
And when streaming became the dominant metaphor for media delivery, Lang says, “that whole momentum that Real created really pushed people off our idea, which was bursting.”
The technical differences between bursting and streaming are complex, but at bottom, which you prefer is suggestive of your worldview. A streamer is essentially an optimist, one who believes the Internet is reliable enough to broadcast, in a steady “stream,” the thousands of bits of information that compose a video and to get them to their destination relatively intact. When you stream data, you send it in real time: A one-hour video takes an hour to send. By doing this, says Lang, you’re implicitly making an assumption about the network. You’re deciding that the network works and that it will work for the full duration of the show you’re sending.
A burster is a pessimist. Lang doesn’t trust the Internet; he sees it as inherently unreliable. You can’t count on its being up and at full capacity for the full duration of your program, he says, and “so what we did was say, when there’s more bandwidth” — when the network is not busy — “you deliver more.” You don’t wait the full hour to send the show but, instead, send it “faster than real time,” transmitting the whole video in the few minutes that the network might be working well.
But the streaming metaphor was more attractive than the bursting metaphor, says Richard Doherty, the director of the Envisioneering Group, a technology research firm. “Everyone loved the word ‘stream’ because it sounded smooth and dependable,” he says. “It’s easy to understand. Nobody wanted to use the word ‘bursty’ media, ‘machine-gun’ media. People wanted to stream.”
The first fight in the development of media delivery on the Internet was over the desktop: Which company’s player would become the one that Web surfers first turn to when they download a video from the Internet? RealNetworks took the early lead in that battle, but Microsoft, with its Windows monopoly and mountains of cash, was not far behind. According to the generally accepted history of how the industry developed, the two firms, whose headquarters are just miles from each other, were soon attempting to pound each other out of existence.
Lang disputes that history. Far from being arch rivals, Microsoft and RealNetworks had an understanding with each other, Lang contends — they would compete, but they wouldn’t play too rough. The companies would divide the market, and small companies like Burst would be kept out. This deal, Lang alleges, allowed Microsoft to kill off Burst, causing Burst to beg Microsoft to license the firm’s technology. And when Burst opened up its technology to Microsoft’s executives, Microsoft took very close notes — and, Lang says, the burst technology eventually ended up in Microsoft’s software.
RealNetworks showed up on Microsoft’s radar in the late 1990s, when Real’s head start in streaming appeared to be giving it a lead in multimedia that Microsoft didn’t appreciate. According to the findings of fact issued by the federal district court in D.C. as part of the Justice Department’s antitrust trial against Microsoft, Microsoft saw RealNetworks as a threat to its way of doing things.
“At the end of May 1997, [Bill] Gates told a group of Microsoft executives that multimedia streaming represented strategic ground that Microsoft needed to capture,” those findings state. “[Gates] identified RealNetworks as the adversary and authorized the payment of up to $65 million for a streaming software company in order to accelerate Microsoft’s effort to seize control of streaming standards. Two weeks later, Microsoft signed a letter of intent for the acquisition of a streaming media company called VXtreme.” (The complete findings, which run a breezy 103 pages and from which Burst.com quotes liberally in its complaint, are available here.)
“At this time,” Lang says, “the battle for the desktop was on.” But his company’s strategy for creating its own desktop player quickly fizzled. Burst had built its media player on the Java Media Framework, which Intel had been working with Sun to develop. The JMF, as it’s called, was to be a cross-platform player that would play a variety of media formats — a prospect that Microsoft saw as a threat to the Windows platform. Intel’s support for the player was frequently described as a being a sore point between Microsoft and the chipmaker, and in 1997, according to the findings of fact, Microsoft and Intel entered into a deal that led Intel to drop support for the system.
It’s probably a stretch to say that Microsoft wanted Intel to drop support for the JMF as a way of specifically hurting Instant Video Technologies. (According to the findings, it was a way of hurting Sun, the main developer of the Java programming language.) It’s not clear that Microsoft even knew of Lang or his company at the time, but Intel’s move had the effect of crippling Lang’s firm anyway. Burst had worked closely with Intel’s Java developers, and now that they were “dropping us,” as Lang puts it, Burst saw no choice but to abandon its dream of creating a media player. Instead, Lang says, the firm now focused on creating a plug-in to let bursting work on other players.
And for a while, the plug-in idea worked well, Lang says. The plug-in that Burst created for Microsoft’s Windows Media player enabled Windows users to get video that looked better than that delivered with Windows Media alone, and it transferred video more efficiently over the network, saving about 25 percent of bandwidth over streaming video.
Lang says that Burst’s technology had become good enough that U2 decided to offer its summer 2000 PopMart concerts for free on the Web using Burst’s software. In the first six months of that year, Burst made $500,000 in software sales, and it raised $12 million in investments, including $5 million from the phone company SBC. Even Microsoft was taken with the company’s technology — it invited Burst to participate in Microsoft’s booth at the National Association of Broadcasters show that year in Las Vegas.
And then suddenly, Lang says, Redmond soured on Burst. Shortly before the U2′s concert tour was to begin, Microsoft released Windows Media Player 7 — and, to Lang’s horror, Burst’s engineers discovered that the new player didn’t support Burst’s plug-in.
“They made changes to their code that only affected Burst,” Lang says. In its complaint, Burst says that it had “long and frustrating discussions with Microsoft to establish workarounds or other fixes, but Microsoft strategically used its power to disadvantage and delay Burst.” As a result, everyone who downloaded the new player couldn’t watch the U2 concerts, and Burst was deprived of the first major “showcase” of its technology.
Was the new Microsoft player’s incompatibility with Burst accidental, or intentional? Was its timing coincidental, or calculated? According to Lang and his attorneys, the changes Microsoft made to the Windows Media Player were so tiny that they seemed designed to stymie Burst: “It was a surgical change,” Lang says. (In its legal response to Burst’s filing, Microsoft explained away this “surgical change” by stating: “As a matter of law, Microsoft has no obligation to deprive its customers of technological improvements in order to help its competitor Burst.”)
Bruce Wecker, one of Burst’s attorneys, concedes that not all of the facts are available yet to prove that what Microsoft did was deliberate. “Because we haven’t had discovery in the case yet, we’re at a phase where we don’t know all they did,” he said. If the case comes to trial, said George Frost, another attorney in the case, “we can put their engineers under oath.”
End of Part 1. Tomorrow, in Part 2, Burst.com shows Microsoft its “special sauce” under a nondisclosure agreement. Less than a year later, Microsoft’s streaming technology looks a lot like Burst’s. Read Part II.
Farhad Manjoo is a Salon staff writer and the author of True Enough: Learning to Live in a Post-Fact Society.More Farhad Manjoo.
Elliott and the friends with whom he recorded in middle school in Texas (photo courtesy of Dan Pickering)
Heatmiser publicity shot (L-R: Tony Lash, Brandt Peterson, Neil Gust, Elliott Smith) (photo courtesy of JJ Gonson photography)
Elliott and JJ Gonson (photo courtesy of JJ Gonson photography)
"Stray" 7-inch, Cavity Search Records (photo courtesy of JJ Gonson photography)
Elliott's Hampshire College ID photo, 1987
Elliott with "Le Domino," the guitar he used on "Roman Candle" (courtesy of JJ Gonson photography)
Full "Roman Candle" record cover (courtesy of JJ Gonson photography)
Elliott goofing off in Portland (courtesy of JJ Gonson photography)
Heatmiser (L-R: Elliott Smith, Neil Gust, Tony Lash, Brandt Peterson)(courtesy of JJ Gonson photography)
The Greenhouse Sleeve -- Cassette sleeve from Murder of Crows release, 1988, with first appearance of Condor Avenue (photo courtesy of Glynnis Fawkes)