By Wendy Seltzer, James Losey, Tom Glaisyer, and Kara Hadge
On June 20, 2009, nearly 150,000 people witnessed the death of 26-year-old Neda Agha-Soltan, but unlike the Iranians who passed her by in the street, they weren't bystanders to the post-election turmoil in Tehran that claimed her life. They were merely the first of over 600,000 who have since viewed a now-symbolic YouTube video that helped propel the opposition political movement forward in the following days of protest. The democratizing power of the Web lies in video like this one--not just because of its content, but because anyone with an Internet connection can contribute to a global dialogue.
But imagine if the person who shot this video had been unable to post it anonymously or if YouTube viewers had to pay to watch it. If online videos were subject to patent licensing fees, users could be charged per-view to capture those fees. Beyond the ethical dilemma profiteering from a tragic death, video licensing could reduce the democratic nature of free and open Internet content to monetizable media. The funny cat videos would be gone forever (perhaps not the greatest loss), but so too would the movement-inspiring Nedas of the future remain unknown.
And why should anyone have to pay to share video online in the first place, you ask? The Web certainly wasn't founded on a pay-to-play premise. When Sir Tim Berners-Lee developed the "World Wide Web" at CERN, he released his newly-created HTML and HTTP protocol to the world free of charge and unencumbered, helping it to become the standard it is today. The developers of Apache webserver software followed suit, offering open source software that now powers more than half of the world's 200 million-plus websites. Today you can browse the Web from phones and televisions; you can author Web content on a hosted blog or your own server; Tweet and send pictures to Flickr from your mobile phone. All without owing the Web's creator a penny.
To serve the multimedia of today, websites have added dynamic updates and rich media content, but browsers have had to use third party plug-ins or standalone applications, such as Adobe Flash, Microsoft Silverlight, or RealPlayer, to show videos. Now, the next major revision to HTML is set to add native video support, via tags that would allow video to be included in HTML5 without the use of a third-party player. All you'll need will be a browser. But when it includes native video support, will that browser be as free and open as those we can use now? At stake is the nature of the Web; at issue are a patent license pool, competing claims, and a little known Supreme Court case with potentially wide-reaching implications.
As the Web incorporates multimedia, some participants want to control -- and charge for -- its video standards.To make an HTML5 video tag work for everyone, browser makers and web publishers need to agree to support common video formats. And that's where patents come into the picture. Some participants in the online video discussion claim that common video codecs (the coder/decoders that translate video files for viewing) cannot be implemented without infringing their patents.
One codec under popular consideration for use in HTML5 is H.264 (a.k.a. MPEG-4 AVC), already used for an estimated 66% of all online video content, including Hulu and YouTube. Yet H.264 is also claimed to be subject to a pool of patents controlled by MPEG-LA, a limited liability corporation that describes itself as the "world’s leading packager of patent pools for standards and other technology platforms." As the MPEG-LA website explains, they are "neither patent owner nor patent user," but their "goal is to provide a service that brings all parties together....[and to] help make markets for intellectual property that maximize profits for intellectual property owners." MPEG-LA maintains a "pool" of patents claimed to be essential to H.264. MPEG-LA coordinates the fragmented ownership of patents, but that doesn't make the patenting of software any better for the Web.
They have set up a structure so that users can pay a patent tax that is split among all the groups who believe they own part of the set of interlocking patent claims. The MPEG-LA participants say that use of H.264/MPEG-4 AVC without a license infringes their patents. The consortium has tried to win us over by extending a no-cost license for "Internet Video that is free to end users," such as YouTube and Vimeo, through 2015. But content providers that charge for video -- or produce a program or device to watch or produce video -- are asked to pay up. Moreover, the current license offers no exception for programs distributed free of charge, so if Mozilla shipped more than 100,000 copies of a browser with H.264 support included, it might have to pay royalties. Worse, users who wanted to tweak the software to modify their browsers, as Mozilla's open source software licenses permit and encourage, would have to worry about patent infringement.
To avoid this morass, many advocate the use of free codecs instead. Two stand out: the Ogg Theora video format, un-patented and developed independently by Xiph, and the VP8 codec that Google purchased and released under a license that permits modifications to be made to it.
Recently, however, MPEG-LA has raised questions about the patent status of these free codecs, suggesting that they might in fact use patented technology from the H.264 pool. This is where things get tricky from a legal standpoint, because patent infringement does not require directly copying someone else's invention in the way that copyright infringement does. Even independently developed technology can be ruled to infringe if someone else got to the patent office first. Software patents are particularly troubling because their vague claims are often read to cover technology way beyond the original invention. In software engineering, many developers facing the same problems and working from similar background knowledge can reach the same conclusions independently -- and a patent issued to any one of them can throw a wrench into the others' work. One who chooses not to patent is still at risk from others who have. The uncertainty around video formats has hindered their adoption.
Further, patent law is supposed to protect inventions, not discoveries, facts, or mathematical algorithms. But if software is just human-readable expression or applied math, then why, many people ask, is it patentable in the first place? The United States Supreme Court could have answered that question when it decided Bilski et al, a case challenging a "business method" patent as unpatentable subject matter. While the entire Court agreed that the business method in question was unpatentable as an "abstract idea," the Court left us muddy boundaries on method patents, including those on computer software. Where Justice Stevens, in concurrence, would have sharpened the definitional boundaries, the court's "less than pellucid" majority leaves the public, and software developers, in the dark. When HTML5 rolls out native video, will that pass the “machine or transformation” test that earns the patent examiner’s stamp of approval? Or will it, too, fall under the abstract idea category of the business method patent? The future is unclear.
The software industry grew substantially before anyone thought to demand patents for it. The Web flourished through its promise of free interoperability. Limiting the incursion of patents -- by recognizing limits to the scope of patentable subject matter -- can again make the field more free for business and non-commercial development and help us to preserve the open Web. If the Bilski decision supports stronger patents, however, the freedom of online video is under threat. At the very least, we have to hope that HTML5 implementation includes VP8 or Theora alongside H.264. Of course, we'd prefer it to support only the open source versions, but that's probably too much to hope for.
The worst-case scenario does not necessarily spell the end of online video. Web video won't disappear overnight if H.264 is the only standard included in HTML5, but if this happens, it may be only a matter of time before patent pool owners will want to exact a toll for video production and distribution. Even if the cost is small, the impact could be far greater. If you had to make a micropayment every time you watched a video on YouTube, would you pass hours idly clicking from one video to the next? Or would you embed a video in your own website? Perhaps, but even journalism has been wary to make the move to a micropayment model. Whether online video serves the purpose of amusing distraction or democracy itself, we're unwilling to risk its future on the interests of the patent pool.
Wendy Seltzer is a fellow at the Silicon Flat Irons Center, University of Colorado Law School, and an Adjunct Research Scholar for the New America Foundation. James Losey is a Program Associate for the New America Foundation’s Open Technology Initiative. Tom Glaisyer is a Knight Media Policy Fellow and coordinator of the New America Foundation’s Media Policy Initiative. Kara Hadge is a Research Associate for the New America Foundation’s Media Policy Initiative.
Cross-posted at "Dispatches from the Digital Frontier," the blog of our parent program, the Open Technology Initiative.