Earlier this month, representatives of some of the biggest tech companies met in Vancouver to find common ground in a long-running feud. The meeting brought together representatives from Mozilla, Google, Cisco and Microsoft as well as a number of smaller startups, and the goal was to come to an agreement on WebRTC, a technical standard that could one day be key to how we communicate over long distances — be it through teleconferences in the office, video chats with grandma or even future telepresence applications that make us feel connected through always-on video appliances.
But things didn’t go well in Vancouver. The atmosphere in the meeting has been described as testy, the parties involved were too far apart to find common ground, and everyone left disappointed, without a consent and any clear path forward. So what does the future hold for WebRTC, and why does it matter?
What’s WebRTC all about
Everyone loves video calling, but no one likes the hoops you must jump through to make it work. Want to make a Skype call? Then you’ll first have to download and install Skype. Is your contact instead using Google’s Hangouts? Then you might have to download a plugin as well. Other services require you to update your Flash player, and any kind of corporate video conferencing solution will most likely make you download a separate desktop client. Move to your mobile phone or switch from one computer to another, and you’ll have to do the whole dance all over again.
Companies like Google, Microsoft and Cisco would love to make this easier for consumers. Google wants you to be able to use Hangouts in any browser without needing to download any additional software. Microsoft would love to ship a browser-based version of Skype, and Cisco wants to make it easier for remote workers to connect to the expensive, Cisco-made video conferencing solutions in their corporate office. And then there are literally hundreds of startups, hoping that interoperability between browsers and devices will be their ticket to break into the real-time communications market.
That’s why everyone in the industry has high hopes for WebRTC, an emerging technology that promises to make video chat (and more) available in the browser, once and for all doing away with the need to download any kind of plugins. Technically, WebRTC is a collection of technologies, including a P2P connection protocol that can be used to initiate video chat sessions, a P2P data exchange protocol that will enable users to chat or transfer files without the need for a third-party server, and a way to access a user’s microphone and video camera. (This is obviously just a high-level overview. For a 40 minute deep dive, check out this video.)
It all comes down to H.264 vs. VP8
Proponents of WebRTC are now pushing to turn it into a common standard that would be adopted by all the major browser makers. Much of this work happens within the Internet Engineering Task Force’s (IETF) RTCWeb working group, and that’s where things fell apart during the IETF meeting in Vancouver earlier this month.
That’s because early on in the process, participants in the working group agreed that there should be a mandatory video codec for WebRTC, meaning that the same video format would be implemented by everyone who was going to use the new standard – not necessarily to use it every time someone does a video chat, but to always have one codec to fall back to that works with every browser and every device.
However, the group could never agree on what that codec should be. Google, which has been one of the biggest drivers behind WebRTC, wants to use VP8 — a codec the company acquired when it bought the codec specialist On2 Technologies in 2009. Google open sourced VP8 in 2010, permitting everyone to use the codec for their own projects without the need to pay any licensing fees.
The move was meant to offer a viable alternative to H.264, the predominant video codec that’s been used in Flash and by many online video services. The problem with H.264 is, at least in the opinion of open source advocates, that it is proprietary. A number of companies, including Apple, Ericsson and Nokia, hold patents that cover parts of H.264, and the patent pool outlet MPEG LA has been asking companies that use H.264 in a commercial fashion to pay licensing fees.
However, proponents of H.264 long alleged that VP8 was also susceptible to patent infringement claims. Google and its allies initially argued that this wasn’t the case, but the mere threat of lawsuits from patent holders slowed down acceptance and deployment of the codec. That’s why Google eventually swallowed its pride, and struck a deal with MPEG LA to shield the codec from any such lawsuits.
Cisco won over Mozilla, but not Google
Only, that wasn’t enough for some of the participants of the IETF’s WebRTC working group. Cisco, for example, continued to push for H.264, because the company has sold a ton of expensive video conferencing equipment to enterprise customers. This equipment is based on H.264, and the company wants it to continue to work as the world is transitioning to WebRTC.
That’s why Cisco made an eleventh-hour push for H.264 a mere week before the meeting in Vancouver. The company announced that it was going to develop a binary component with its own implementation of H.264 and freely distribute it to anyone who’d want to take advantage of it. Think of it as a kind of plugin, minus the complicated installation. And in a kind of coup, Cisco won Mozilla, long a proponent of VP8, as a first major partner to make use of its H.264 implementation. At the time, Cisco’s Collaboration CTO Jonathan Rosenberg told me: “We think this will help to push the edge over to H.264.”
Except it didn’t, at least not in Vancouver. Google rejected the proposal publicly and within the IETF working group, arguing that VP8 is still the better solution because it’s an open codec. After the Vancouver meeting ended without an agreement, Rosenberg sent me the following statement:
“We’re disappointed that the IETF failed to reach consensus. We plan on fully continuing our initiatives in making H.264 available and working with Mozilla in enabling H.264 support in Firefox. In parallel, we will continue to participate in the IETF processes in the hopes that consensus can reached in the near future.”
Mozilla CTO Brendan Eich added via email:
“We hope the IETF resolves this quickly and we intend to move forward with our announced plans.”
It’s not chess, it’s poker
Rosenberg also suggested in a blog post Monday that there still may be some wiggle room to find a consensus. But the continued stalemate is now starting to wear on all participants, with some members pleading on the working groups email list that it’s time to move on, and actually work on some of the technical issues of WebRTC. Except, no one really knows how the IETF could resolve this quickly, save for one of the sides giving in — and there’s no indication that this will happen any time soon.
That’s why some are exploring alternatives. One would be to go back and revert the earlier decision that there should be a mandatory codec at all. Why not just leave it up to each party to use the codecs they want? The answer is obvious: Because that might lead to a world in which Google only uses VP8 while Apple or Microsoft only use H.264. To make their browsers speak to each other, services would have to transcode. There is a whole cottage industry around transcoding that would like nothing more than such an outcome, but many others are objecting to that idea.
Others have proposed that the working group should instead settle on an older codec that is less controversial. It could be one of those compromises that makes no one happy but that everyone can live with – except, once you start going down that route, it turns out that none of the other codecs are really free of controversy either. Because in the end, it’s not about technical arguments, but about politics.
And as such, it might just require some grand gesture or other kind of intervention. At least that’s what Monty Montgomery thinks. Montgomery has been a long-time open codec advocate and is in fact currently working on a next-generation video codec for Mozilla, but he recently admitted that the advocates of open codecs may have lost this battle. Asked how the current stalemate around WebRTC could be solved, he told me:
“If the stalemate gets broken decisively, it will probably be something we don’t expect, or at least, something we don’t expect to make the difference. That’s the thing about corporate politics; most of the cards are hidden till they get played. It’s not chess, it’s poker.”