Photograph by Rory McNamara
Giant Slayer: Burst.com CEO Richard Lang, seen here in the darker moments before Microsoft made a $60 million settlement with his firm.
Paradigm Lost, Paradigm Found
Santa Rosa streaming media company Burst.com battles the Microsoft octopus and comes out on top
By R. V. Scheide
Microsoft wants to be in your face–everybody’s face. Capturing more than 90 percent of the desktop operating system market with Windows is evidently not enough. Founder, chairman of the board and chief software architect Bill Gates–whose own net worth, if all things hold equal, will exceed $1 trillion by 2019, according to some calculations–appears to settle for nothing less than total domination of the technology sector. Windows is the octopus that Microsoft executives use to “embrace, extend and extinguish” all other technological life forms that get in its way.
When Netscape debuted the first Web browser available to the public in 1994, that octopus was caught napping. Once awakened, however, Microsoft quickly moved to smother its nascent competitor, extending its Explorer browser into the Windows operating system and seizing the Internet space as its own.
And when Progressive Networks (now known as RealNetworks) launched the first digital media player in 1996, Microsoft once again slithered onto the scene, its far-ranging tentacles embracing, extending and slowly attempting to extinguish Real and other perceived threats in the streaming-media space. The release of Corona in December 2001, the software juggernaut’s “third generation streaming media platform,” might have put the last, deadly squeeze on the competition.
However, there was one slight hang-up. As the details of Corona (later renamed Window Media Series 9) were revealed in the coming weeks, they sounded all too familiar to Richard Lang, CEO and founder of Burst.com, a streaming media company. That’s because Lang and his colleagues had spent the better part of the previous two years sharing their streaming media technology with Microsoft in order to persuade it to license Burst’s software. Microsoft signed a nondisclosure agreement not to reveal any of Burst’s trade secrets, but after 18 months of meetings, it abruptly broke off licensing negotiations.
Then, Lang claims, it hijacked Burst’s technology.
“All of the new and distinguishing features that define Corona are Burst technology,” Lang alleges from his company’s small but tidy downtown Santa Rosa office. Here, he and Burst’s two remaining employees have waited as their lawsuit charging Microsoft with antitrust violations, breach of contract, restraint of trade and patent infringement grinds through federal court. Like the moray eel, the natural enemy of the octopus, Burst took a bite and refused to let go. Last week, its tenacity paid off.
Faced with a pretrial hearing to determine whether it had purposely destroyed e-mails relating to not only the Burst case, but to litigations stretching back to 1994, Microsoft agreed on Thursday, March 10, to license Burst’s patents for a one-time fee of $60 million.
For Lang, 51, a compact, wiry man with streaks of white in his close-cropped dark hair and the focused intensity of a Formula One driver, it’s been one heck of a ride. Since 1987, he’s been perfecting and patenting his own distinct streaming-media system–even way back then, he was calling it a “new paradigm”–and now his company’s hard work is about to pay off. The exact details of the settlement remain to be worked out, but Lang can’t help but feel vindicated.
“The reality is, it’s just a rounding error to them,” he says. “We’re like a plumber that comes to your house to fix your pipes. All we want is to get paid.”
When used in a technical or scientific setting, the word “paradigm” refers to the accepted theoretical framework of a given process. For example, back in the 1970s, almost all telephones plugged into the wall. With the advent of cell phones came a new paradigm.
To understand what Richard Lang means when he talks about a new paradigm in digital audio and video delivery, plumbing provides as apt an analogy as any. It’s common nowadays to refer to one’s connection to the Internet as a “pipe.” Data flows like water from Internet providers to clients through this pipe. The bigger the opening in the pipe, the more data can move through it.
But back in the 1980s, when Lang first conceived of his new paradigm, the Internet as it is known today was still mostly a pipe dream. Lang worked in the world of videotape, where he co-patented the first dual-deck video cassette player in 1983. Four years later, he met Lisa Walters, his future wife, and they went into the video production and distribution business. They originally planned to set up kiosks to sell children’s and documentary videos in bookstores across the country, but quickly ran into a problem: no one was very interested in providing precious floor space to hawk somebody else’s product.
By the late 1980s, Lang says it had become clear that advances in digital technology would soon render magnetic videotape obsolete anyway. While no one at the time was certain exactly how digital video would be distributed, broadcasting was the dominant paradigm. The idea most people were exploring, Lang says, “was to extend TV over the Internet.” But he had a different idea.”I wasn’t thinking of playing it over a network like a TV show,” he says. “Because I came from the world of VCRs and video cassettes, I wanted to deliver the video over the network like a package, and I devised the apparatus and the method to do that.”
Lang, who calls himself an “inventrepreneur,” is a classic ideas man. He has a communications degree and a foreign-language minor, but doesn’t know a lick of computer code. “You don’t have to know code to be able to implement ideas,” he says. Instead, he utilized the expertise of two groups, software engineers and patent attorneys, to develop his concept.
In the old broadcast paradigm, digitized video and audio files flow through the Internet pipe in “real time” in much the same way that television and radio are coded, transmitted over airwaves, and then decoded by a radio or a television receiver. For example, a 30-minute video clip “streams” from a special server through the pipe for 30 minutes, attempting to keep a small storage space on the desktop known as the “buffer” filled so the end user can simultaneously view the video on his or her media player in real time.
However, computer networks are not as reliable as airwaves. Excess network activity, inherently large video and audio files, or both in tandem can clog the pipe. Such interference typically occurs at peak usage times, but to a great degree, the amount of space available in the pipe continually fluctuates, which in the “real time streaming” paradigm, ends up starving the buffer of content. This leads to the herky-jerky quality evidenced in the first digital songs and videos made publicly available via the Internet and still visible today for those who lack a high-speed cable Internet connection–a so-called fat pipe–or a delivery technology that incorporates Burst’s innovation.
In Lang’s new paradigm, time is turned upside down and inside out to address the problem. By eschewing the old broadcast strategy of streaming in real time, Lang’s software engineers asserted that files could be sent through the pipe “faster than real time,” a phrase Lang trademarked in 1991. Using this method, a 30-minute video can be “burst” (as opposed to “streamed”) through the pipe in a matter of seconds, either all at once or in chunks, where, taking advantage of the ever-increasing storage capacity of modern PCs or set-top boxes, it is stored to begin playback.
Robert X. Cringley, a noted computer consultant, technology writer and host of the PBS miniseries Electric Money, explains that such high speeds are possible because Burst’s software searches “for headroom in the pipe, trying to keep the pipe full 100 percent of the time.” Unlike streaming digital media, the file isn’t downloaded in chronological order–the sequence in which it is viewed or listened to. Instead, it’s downloaded according to the size of the data bits (a romance scene, Cringley suggests, might be smaller than, say, an action scene) and the amount of available headroom in the pipe. “Bursting sends different parts ahead of time. It takes a lot less time.”
Once a large enough portion of the file is loaded in the client’s media player, Burst’s technology takes advantage of the increased computing power of modern PCs, and begins playing the video in real time, drawing on the program stored locally on the PC or set-top box instead of continually sapping the network. Because the entire chunk is downloaded in a fraction of the time it takes to watch it, bandwidth–the amount of room available in the pipe–is conserved and the clip is isolated from such nuisances as clogged networks and slow connection speeds. The viewer experiences a smoother, broadcast-quality experience.This is an admittedly gross simplification of a process for which Burst, beginning in 1990 under the name of Explore Technology, has received nine U.S. patents and 25 international patents. There are some in the video-streaming industry–most notably Microsoft–who question whether Burst has developed any independent innovations at all. But way back in 1990, there seems to be no question that Lang’s company had a major innovation on their hands.
A chance meeting in 1989 with the Irish rockers U2 led to a $2 million investment in Explore Technology by the band. In these early days, Lang’s company devised and built its own hardware, the instant video transceiver and receiver, to demonstrate its instant video technology. The devices were the buzz of the 1991 Winter Consumer Electronics Show in Las Vegas. “[T]he technology has the capacity to revolutionize the transmission and reception of programming for broadcast and cable operators,” the Christian Science Monitor excitedly reported.
For the first half of the 1990s, Explore Technology focused on developing patents and technology and attempting to convince the industry, through conventions and trade shows, that its new paradigm was the way to go. In 1992 it changed its name to IVT (Instant Video Technologies). “We spent much of the 1990s beating people over the head with it,” Lang says. Then in 1996, the debut of RealNetworks’ (then known as Progressive Networks) Real Player signaled that the paradigm shift had begun in earnest.
“For us, that was the alarm bell,” Lang says. RealNetworks method of delivering audio and video wasn’t as sophisticated or efficient as Lang’s, but they had a product on the market, ready to go. His company shifted from its earlier plans to license and manufacture hardware and changed its name to Burst. “It was time to implement our technology as software.” Burst.com immediately began collecting a low-profile stable of high-profile clients and investors, such as SBC, AOL and [email protected] “By 2000 we were about to really take off,” Lang says. Burst had grown to 110 employees, with offices in the heart of San Francisco’s financial district. “SBC invested $5 million and signed an agreement to use Burstware in its DSL networks,” Land says. “Burst in every DSL home.”
Burst had only two major competitors, Real and Microsoft, headquartered just miles apart in Redmond, Wash. Although Apple had debuted a version of its QuickTime movie player as early as 1991, it wasn’t viewed as a direct threat because of Apple’s limited share of the operating-system market. Microsoft, however, with its monopoly of the operating-systems market (Windows was used on 96 percent of all PCs last year, according to the New York Times), could not be ignored.
“We knew that for our products to work properly, they had to interface with Microsoft,” Lang says. In 1999, hoping the software giant would license its technology, Burst entered into an agreement to share its trade secrets and technology with Microsoft. “We thought, ‘We’ll protect ourselves the best way we can.’ We got them to sign the special nondisclosure agreement.”
As it turned out, that wasn’t quite enough.
Streaming Snake Oil?
The tricky thing about paradigm shifts is that unless you were on what was called the “bleeding edge” in the 1990s, it’s difficult to detect when the old paradigm ended and the new one began. By the late ’90s, Burst as well as most of its competitors were using codecs–the mathematical algorithms programmers use to compress and decompress digital files–in their streaming-media products. But according to Burst’s patents, compression was just one component in a larger design of methods and apparatus that Burst had developed. Some in the streaming-media industry question the validity of Burst’s claim to the new paradigm.
Charles Wiltgen, a computer consultant and Apple’s QuickTime “evangelist” during the 1990s, called Burst’s claim to the new paradigm “streaming snake oil” in a scathing column in 2002. Responding to an article in Salon.com documenting Burst’s lawsuit against Microsoft, Wiltgen wrote, “While I was Apple’s QuickTime Evangelist, I was a magnet for all kinds of folks who claimed to have miraculous codecs and other holy-grail technologies. Burst.com claimed to have a revolutionary way of delivering streaming content. Lossless. Faster than real time. Well, golly. You can deliver content losslessly and faster than real time via HTTP and FTP, too. Only Burst.com did this with a magical, proprietary protocol that required a magical, proprietary server that they would be happy to sell to you.”
Wiltgen, responding by e-mail, says he still stands by the comment. However, Robert X. Cringley disagrees with this assessment of Burst’s product.
“In 1991 their technology was innovative,” he says. “It wasn’t just an idea, it was an implementation. Now it’s commonly done because their patents are being infringed upon. Both Apple and Real are looking at Burst’s Microsoft suit very closely.”
At any rate, Lang says it’s common for people in the industry to not understand Burst’s innovations, in large part because the distinctions were not obvious, a key element of patentability. That was certainly his company’s experience during its initial meetings with Microsoft.
“At our first meeting in 1999, they scoffed at what we had,” he says. “They didn’t understand it. They had just invested hundreds of millions in edge-caching technology, primarily designed for static web pages. They said, ‘Give us more proof.'”
Burst engineers met with members of the Windows Media team seven times over the following 18 months, both at Microsoft’s Redmond campus and at various trade shows. A Burst-enabled version of Windows Media Player was created, and when tested by an independent lab routinely used by Microsoft, proved to be highly more efficient than the standard media player, Lang says. For the Windows Media team, the light was beginning to flicker on–but Microsoft still wanted more information.
“We had secret trials going on, and we put them in touch with our customers and our large potential customers, like SBC,” Lang says. “That we even had such large potential customers was a trade secret. We put Microsoft in touch with them and they went out and talked to them.”
As the meetings with Microsoft were taking place, Lang was also working hard preparing for Burst.com’s public launch: a June 2000, video-on-demand Webcast of U2’s PopMart concert tour from Las Vegas, the first of its kind. Burst was now at full strength with 110 employees, and Lang spent $1.5 million building a hosting network and promoting the concert. Then disaster struck.
“Three weeks before the launch date, Microsoft announces a new version of its player,” Lang recalls. “And guess what? There is only one existing software that doesn’t work on it: Burstware. They broke our player, right before its public launch. They pretended to help us, but the problem was never fixed.”
Windows users–which is to say the vast majority of PC owners–eagerly downloaded the new player, only to discover they couldn’t watch the concert. Burst.com’s downward spiral had begun.
“All of a sudden, our customers stopped talking to us,” Lang says. “They just walked away. Our investment banker just walked away. I’d go home to Lisa and say we’d just lost another customer.”
“There must be something you’re missing,” his wife told him. Lang would not discover the full extent of just what he was missing until the discovery process of its lawsuit against Microsoft was completed in 2004.
Meanwhile, in January 2001, Microsoft offered Burst.com “up to $1 million” to license Burstware, a paltry sum, considering the time and money Lang’s company had invested in the project. Lang made a counteroffer. Microsoft turned it down and broke off communications. By March, Burst had been downsized to just five employees. By the end of the year, they were down to two. The coup de grâce came in December, 2001, when Microsoft publicly released Corona with, Lang alleges, many of the same features of Burstware.
Burst.com hit rock bottom.
Embrace, Extend, Extinguish
When Lang first approached Microsoft about Corona’s similarities to Burstware, he says the company informed him that it hadn’t “taken anything that was secret.”
“Once the paradigm shifts, then it’s obvious to everybody,” Lang explains, a dozen or so neatly framed patents hanging on the office wall behind him. “Then everybody goes, ‘Why should you get the money? It’s obvious.’ That ignores 10 years of work we had influencing the shift.”
After numerous calls from members of the streaming-media industry, journalists and others who noted Corona’s similarities to Burstware, Lang decided to file suit. Two San Francisco law firms agreed to take the case on contingency: Hosie, Frost, Large & McArthur, which has extensive experience in antitrust and unfair trade practice; and Carr-Ferrell, with expertise in intellectual-property litigation. The suit, filed in June 2002, contends that Microsoft wrongfully stole Burst’s technology and used its Windows operating system–legally declared a monopoly by federal judge Thomas Penfield Jackson in 1999, a decision upheld upon appeal–to “embrace, extend and extinguish” Burst.
“‘Embrace, extend, extinguish’ is a phrase key Microsoft executives use repeatedly to do their business,” says Spencer Hosie, Burst’s lead attorney on the case. In industry parlance, Microsoft’s practices have come to be known as “Netscaping,” in reference to the Mountain View company that released the first publicly available Web browser in 1994 and subsequently came under concerted Microsoft attack. AOL/Time-Warner, now Netscape’s parent corporation, was eventually awarded $750 million in damages when Microsoft settled out of court last year.
As Burst’s attorneys learned during the discovery phase of the lawsuit, there are interesting parallels between the Burst and Netscape cases. Netscape had completely blindsided Microsoft, which had failed to anticipate the advent of the Internet, with its browser. Microsoft, perceiving the browser as a potential threat to its operating-system monopoly, proceeded to enthusiastically embrace, extend and extinguish Netscape. Likewise, documents obtained by Burst’s attorneys revealed that in 1997 chairman Bill Gates was concerned that RealNetworks presented a similar threat to Microsoft’s dominance. RealNetworks, one Microsoft executive said, “is like Netscape. The only difference is we have a chance to start this battle earlier in the game.”
Under pressure from Microsoft, RealNetworks agreed not to compete directly on the fundamentals of streaming audio and video. In exchange, Microsoft agreed to support “RealNetworks as a value-added software provider.” Microsoft would own the streaming-media platform, which would be incorporated into its operating system. However, RealNetworks never honored the deal, one which potentially violates antitrust law.
“They took Microsoft’s money, then said ‘Fuck you’ as soon as the check cleared,” Hosie says. “Microsoft can’t sue Real for failing to honor an illegal agreement.” In December 2003, Real, facing financial difficulties like Burst, issued Microsoft another raspberry, suing the company for “predatory action over a period of years by abusing its monopoly power, resulting in substantial lost revenue and business for RealNetworks.”
Evidence uncovered in discovery seems to indicate Microsoft pursued a similar strategy with Burst. As Burst was imploding under the twin pressures of the dotcom bust and Microsoft’s alleged anticompetitive practices, an e-mail exchange between members of the Windows Media team reads, “Check out their stock price. Going, going . . .” Furthermore, Lang alleges, in its patent applications for Windows Media Series 9, “they essentially duplicated what our product did and what we showed them. The very same people we worked with are named as inventors.”
But potentially far more damaging to Microsoft is what Burst’s lawyers didn’t find in the thousands of documents dumped on them by the software juggernaut. Despite meeting with members of the Windows Media team seven times over 18 months, there were virtually no e-mails, internal or external, concerning the meetings. “Each time we had a meeting, there are gaps in Microsoft’s e-mail record,” Lang says. “After the very first meeting, they sent an announcement to the entire media division. Supposedly, not one reply to the document was received.”
What Burst’s attorneys discovered was that Microsoft has prohibited employees from saving e-mail for more than 30 days on its corporate servers since 1994. Even in situations where Microsoft was ordered to retain records related to ongoing litigation, it appears to have been very selective, for its own benefit, as to which employees it ordered to retain e-mails. In a pretrial motion filed last October, Burst alleged that by doing so, Microsoft has destroyed evidence not only in Burst’s case, but in many of the cases which have already been tried, including the Department of Justice case which resulted in Judge Jackson declaring Microsoft a monopoly.
In that case, the court repeatedly requested all documents pertaining to negotiations between RealNetworks and Microsoft, as well as communications between Intel and Microsoft regarding Intel’s Java media framework (JMF). No documents were ever received by the court, thanks, Burst alleges, to Microsoft’s e-mail retention policies. The Intel communications are important because Burst had originally designed its own media player to run on JMF, which works across different operating systems. Lang alleges that Microsoft coerced Intel into dropping JMF because the product was “giving great momentum” to Java programming language developer Sun Microsystems–perceived by Microsoft as yet another threat to its Windows monopoly. (Last year, Sun received a settlement from Microsoft estimated to be worth as much as $2 billion.) When Intel dropped JMF in 1997, it effectively killed Burst’s independent media player, leaving the company with little choice but to design a player plug-in for Microsoft’s Media Player.
In its spoliation motion, Burst requested that when the case goes to trial, the jury be instructed that the e-mails were intentionally destroyed by Microsoft with the assumption that the information destroyed would have hurt Microsoft and helped Burst.
On Thursday, March 10, Judge H. Frederick Motz was set to rule on the spoliation motion when Microsoft blinked.
When the octopus wraps its tentacles around the moray eel, the eel ties a knot in its tail and slides the knot up its body to break the creature’s grasp. By convincing Microsoft to license Burstware for a one-time fee of $60 million, that’s exactly what Burst.com has done.
“They’re still an octopus,” Hosie says. “But with one slightly smaller tentacle.”
Did the Octopus blink because of fear that once Microsoft’s e-mail retention polices were made public, a whole slew of new litigation might ensue? Not exactly, according to Microsoft spokeswoman Stacy Drake.
“We were very confident about prevailing in the case, but we wanted to resolve it without the risks of litigation,” she says. Still, Drake denies that there are “any similarities between Microsoft and Burst’s technology.”
That’s OK with Hosie.
“From our perspective, once Microsoft did the responsible thing and licensed my client’s technology, we really have inner peace with Microsoft,” he says. “They did the right thing.”
The proposed settlement caps off two years in which Microsoft has paid a total of more than $3 billion to settle cases with companies such as Sun Microsystems, Time Warner Inc. and Novell Inc.
“I think if you look at our efforts over the past two years, you’ll see a pattern,” Drake says. “We’re solving the conflicts of the past and moving forward, to develop new products for our customers in the future.”
Burst, meanwhile, moves on to another new paradigm. As Robert X. Cringley pointed out in his New Year’s prediction column for PBS.org, the streaming-media industry has been watching the case closely–particularly RealNetworks, Apple Computer and set-top box manufacturers like TiVo.
“The license from Burst’s perspective is powerful validation of the integrity of the Burst intellectual-property portfolio,” says Hosie. “Microsoft paid a one-time licensing fee of $60 million. That’s essentially Act I, Burst. Act II begins next week. Burst is going to enforce its intellectual property against a series of companies that are currently infringing on their patents.”
“They know who they are,” says Lang.
From the March 16-22, 2005 issue of the North Bay Bohemian.