"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)
[Read Part 1.]
If Burst.com CEO Richard Lang is right, and his company’s video-delivery technology was better than everyone else’s, then two questions must be addressed. Microsoft had identified the delivery of digital entertainment as a key market, and it had a formidable foe to beat in that market: RealNetworks. If Microsoft had hoped to catch up to RealNetworks, why wouldn’t it have wanted a plug-in that improved its technology, instead of making the new version of its media player incompatible with Burst? And why wasn’t it afraid that RealNetworks would get access to Burst’s technology instead?
Burst answers those questions with two of its most damaging charges against Microsoft. Lang says that at first Microsoft wasn’t fully aware of what made Burst so good. So in late 1999, under a nondisclosure agreement and as part of an effort to get Microsoft to license Burst technology, Burst showed Microsoft the company’s “secret sauce,” Lang says. The company disclosed technical information that went beyond what was published in its patents.
At least one of those meetings, Lang says, was with Will Poole, an influential Microsoft executive who directed the company’s efforts in digital media. “And we could see the light bulbs go off,” Lang says of the meeting. “We could see they got it. They understood why Burst was better.”
But Microsoft declined to license Burst’s technology at that time. Again why would it make that decision, if it was afraid of competition from Real? Lang’s most explosive allegation is that Microsoft was in fact not afraid of competition from Real, because the two companies had long ago agreed to divvy up the market. On July 18, 1997, RealNetworks and Microsoft signed an agreement that, according to Burst, was something of a Faustian bargain for RealNetworks: In it, according to legal filings made by Burst, Microsoft essentially agreed not to “target” Real if Real gave Microsoft “veto power” over RealNetworks’ software development, including Real’s ability to deal with other companies.
According to this theory, then, Microsoft was never afraid that Burst’s technology could go to anyone else, because it controlled what Real did. Therefore, it had no incentive to license from Burst.
By the end of 2000, Burst was a company in trouble. “The Nasdaq had begun to implode,” says Lang, “and our investors were asking if we were ever going to license to Microsoft or RealNetworks. And it didn’t make sense to us or our investors. We knew our technology was quality, but we couldn’t see why one of them wouldn’t want to sign with us.” Lang reduced the company from 100 workers to five.
Burst continued to pursue a licensing deal with Microsoft, and in January 2001, after the firm showed Microsoft benchmark tests done by Approach, a technology company, which proved that Burst was more efficient than Windows Media, Microsoft said it wanted to license Burst’s technology. It offered Burst $1 million for all rights “exclusively and in perpetuity” to Burst’s software — an offer Lang considered “a joke.” He made a counteroffer, which Microsoft rejected. Negotiations ended.
“They knew we had nowhere else to go,” Lang says, “and they left us on the vine to die.”
The actual agreement between Microsoft and RealNetworks is secret, so it’s unclear what sort of relationship Microsoft had with Real. News reports at the time painted the deal as nothing more than a routine Microsoft investment in interesting technology: Microsoft would gain a 10 percent “nonvoting” stake in the company, and Real would try to push Microsoft’s streaming format in its player.
Was the deal between Microsoft and Real anything more than that — was it “collusion”? That charge will perhaps be the most difficult, and the most important, for Burst to prove. That’s because, if the Burst case ever goes to trial, Real and Microsoft’s “conspiracy” to keep others out of the market for video streaming will be the only evidence that Microsoft had a motive to appropriate the technology, as Burst says it did, after Burst showed Microsoft the technology under a nondisclosure agreement.
In its one filing in this case — a motion to dismiss the case based on what it suggests are Burst’s flimsy antitrust claims — Microsoft says that Burst is offering an incorrect reading of the contract it signed with RealNetworks. “The contract was manifestly pro-competitive,” Microsoft says, and it “simply does not contain any provision that prevented RealNetworks from dealing with Burst.” There was also no provision, Microsoft says, that prevented Microsoft and Real from aggressively competing with each other — meaning that if Burst’s technology really was as good as it says it was, either company was free to license it.
“If there was collusion,” says Jim Dessler, a Microsoft spokesman, “that would mean there would be no innovation going on, and innovative companies such as Burst and others would have had advantages.” But he suggested that was not the case as, during the past few years, Microsoft and Real competed aggressively against each other, and each “innovated” in the video market.
Dave Richards, RealNetworks’ vice president of consumer systems, declined to discuss the Burst.com case, but he listed a number of improvements that RealNetworks had made to its system after the first release, and he said the firm did consider Microsoft a competitor “in certain areas.”
Richard Doherty, who follows media development at the Envisioneering Group, also finds the accusation that there was collusion between Microsoft and Real difficult to believe: “I find that hard to justify. I would say that on both the technology and marketing side, that one-upsmanship has been something that everyone has enjoyed. Microsoft and Real have each out-innovated each other more than anyone else could have come close to. And if you look at the people involved on both sides, I don’t see that agreement coming from them: from Will Poole and Bill Gates to Rob Glaser and his tech team. If you met Rob Glaser, you’d know that the last thing he’d be involved in is some sort of conspiratorial gentlemen’s agreement with Gates.”
But Lang says he has no doubt that Microsoft and Real worked together to keep others out of the market. He says that the antitrust trial — in which a RealNetworks executive testified that he remembered the “cashectomy” Microsoft had performed on Borland software and wondered whether RealNetworks could become such a target — indicates that Real knew that Microsoft could hurt the company if it wanted to.
The “findings of fact” portrays an aggressive Microsoft: “Microsoft’s intentions toward RealNetworks in 1997, and its dealings with the company that summer,” it states, “show that decision-makers at Microsoft were willing to invest a large amount of cash and other resources into securing the agreement of other companies” whose work threatened them. (Citing attorney-client privilege, Bruce Wecker, Burst’s lawyer, declined to say why, if Burst was arguing collusion between Real and Microsoft, it wasn’t also suing Real.)
In December of 2001, about a year after Burst showed Microsoft its “secret sauce,” Microsoft announced “Corona,” its next version of the Windows Media products. One of the technologies featured in Corona — now called Windows Media 9 Series — is FastStream, which, according to Microsoft’s first press release on the system, “automatically optimizes the delivery of streaming audio and video to take advantage of the full bandwidth available to the user.”
When Lang heard of this technology, he thought it sounded very similar to what Burst had been doing. He looked further into the system, and as he heard more, he says, he got increasingly suspicious. FastStream, he says now, is Burst’s technology. “I don’t think there’s any question they’re violating our patents,” he said. More than that, Burst alleges that Microsoft incorporated ideas that Burst had given the company that weren’t included in its patents. “Even if they didn’t copy our sauce,” he said, “we showed them the ingredients. They knew what had to be in it. They could say they made their own sauce, but they knew our ingredients.”
It’s true that Microsoft’s FastStream technology does sound very much like Burst’s system. Michael Aldridge, a Windows Media exec, described the technology to me. “In previous streaming technology, both ours and anybody else’s, when you send out a 200 kb/s [kilobits per second] video on a 500 kb/s network, it wasn’t intelligent enough to find out that I had an extra 300 kb/s of headroom,” Aldridge said. The new Windows Media is “able to do faster-than-real-time delivery, so more of the file is residing locally than was capable previously. The simple example I would give you as a rule of thumb is, I’m watching a music video. In the first minute of the five-minute video, I might already have the whole thing residing [on my hard drive.] So that even if I lose my network connection I’d still be able to view my content. That helps to iron out some of the irregularities of the network.”
Aldridge’s description of FastStream, especially his use of the phrase “faster than real time,” is almost identical to Lang’s description of Burst.
But Aldridge also said that Microsoft had been working on this system for four years and that it had spent a lot of resources perfecting the product. And Bill Gates has said that during the last couple of years Microsoft has spent $500 million on Windows Media technologies — “which is a lot, even for Microsoft,” says Matt Rosoff, who tracks the company for the research firm Directions on Microsoft.
Which leads to another question: If Microsoft was spending so much money on digital media anyway, and ended up with a technology that at least sounds like Burst’s, why wouldn’t it have just licensed Burst’s system? What could have been its reason for “stealing” from Burst — and would it have risked litigation to do so?
Microsoft didn’t comment on the specific intellectual-property claims in this case, except to deny them broadly. “I think it’s also important to say, not addressing this case specifically, that you had the bubble in the tech sector which was fueled by unrealistic expectations concerning the value of business and in some cases unrealistic expectations regarding the return on intellectual property,” Dessler said. “Now that the bubble has burst, people are looking at ways to mitigate their losses and we’re seeing a growing number of patent-infringement suits. Unfortunately that’s a part of the business now.” He added that Microsoft is an “intellectual-property company, and we respect and understand the value of those rights in the industry.”
Why would Microsoft have brazenly stolen Burst’s technologies instead of licensing them? “It’s our suspicion that they thought they could get it without having to pay,” says Burst attorney Wecker, referring to the Real deal. In other words, Microsoft had no incentive to license Burst when it knew that Burst would have nowhere else to turn.
But Burst had a portfolio of public patents protecting the technology — wouldn’t Microsoft have seen that as an obstacle?
“Patents are useful,” says Wecker, “but they can be challenged.” In other words, Microsoft may have violated the patents only because it thought it could get away with doing so.
According to Mike Madison, a patent expert at the University of Pittsburgh Law School, patent cases are extremely difficult to prove. A defendant usually argues that a patent is not valid, and then if the firm loses on that claim, it argues that its process doesn’t violate the patent. Proving that a patent has been violated is expensive and time-consuming, which plays to the advantage of big firms with lots of money.
None of this means that Microsoft is guilty of any wrongdoing. Doherty, of Envisioneering, says that regardless of who holds the patent, what may count is the “implementation of those patents.” If you look up the engineering notes at all the firms, he says, “you’ll likely see a lot of parallel discoveries. If you deposed everyone you might find that someone in Burst came up with the idea on March 13, and someone in Microsoft came up with it on March 15. Now, Burst is entitled to feel proud, but what usually happens is that even though someone patents something early, the people who make money are the ones who come out with products closer to market.”
Did Microsoft have an incentive to violate Burst’s patents, though? As a matter of pure speculation, it’s clear the company would have had some cause to not care too much about Burst’s intellectual property. Not only might the company have felt relatively certain that it would win a patent case, but the stakes are in any case huge.
“Microsoft’s strategy is to popularize the Windows Media format,” says Directions on Microsoft analyst Rosoff. By most accounts, Microsoft’s video formats are already better than open standards already in use. They yield smaller files of higher fidelity than MPEG-2, which is what DVDs and MP3s are coded in, and MPEG-4, which many competitors of Microsoft hope will become the “next generation” format for digital video.
Already, many digital music players and DVD players can play Windows Media audio files, and Aldridge, at Microsoft, says that some DVD players might soon play Windows Media video discs in addition to standards-based DVDs. “I can’t imagine people going out en masse tomorrow or even next year to throw out their DVD players to buy new Windows Media DVD players,” says Rosoff, “but Microsoft looks at the long run. And in the long run it may be possible.”
And the first step in making the format popular for devices is making it popular on the Internet. And the key to making it popular on the Internet is make it stream faster than every other player. But Microsoft’s route to that last step, according to Lang and his lawyers, was blocked by Burst’s patents.
In the beginning, says Lang, he trusted Microsoft. “This was during the antitrust trial, but despite what everybody said, I had always given them the benefit of the doubt. And you’ve got to remember they always acted friendly.”
But as things started to fall apart for his company, it became clear to him that Microsoft wasn’t going to let Burst succeed. “And this company — we played by all the rules. We’d had the vision early, way early. We did it all by the rules, and even when we encountered all these roadblocks we still went ahead. And in the end we were right — it was our technology that was the best.”
That should have been enough, he says — but it was not. Which is why he’s now here, at his lawyer’s office. “If there’s such a thing as justice in the business world, we’ll end up recouping the value of our technology,” Lang says as he looks out at the bay. “Litigation was never our business plan.”
Farhad Manjoo is a Salon staff writer and the author of True Enough: Learning to Live in a Post-Fact Society.More Farhad Manjoo.
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)