Blog Post

Microsoft commits to WebRTC – just not Google’s version

Microsoft (s MSFT) officially announced its own proposal for plugin-free real-time communication Monday. “Customizable, Ubiquitous Real Time Communication over the Web,” or short CU-RTC-Web, is Microsoft’s contribution to the W3C WebRTC working group, which is working on a common API for voice and video chat between browsers.

Microsoft’s commitment to this kind of technology is a big deal for the future of Skype and other messaging applications. The company’s Skype unit has been working on a browser-based version of its voice and video calling software, and WebRTC could be key to making Skype accessible on a wide variety of browsers without the need for any applications or plug-ins. The technology could also help Skype open up its walled garden, and enable interoperability between users of Skype and Google (s GOOG) Talk or other third-party services.

I had a chance to talk to Matthew Kaufman, principal architect for Microsoft-Skype on WebRTC, about his company’s support for WebRTC last week. He didn’t want to comment on how exactly Microsoft wants to use WebRTC, but the mere fact that he, as a key Skype executive, is heading these efforts speaks volumes.

Microsoft kept low profile while others were already implementing

Kaufman also told me that both Skype and Microsoft have been closely involved in the work on a WebRTC standard at the IETF and the W3C since 2010. However, until now, this work has been pretty quiet – especially if you compare it to other browser vendors. Google was an early vocal supporter of WebRTC, and the company recently included key parts of the technology in the stable version of Chrome. Both Mozilla and Opera also ship browsers that already support WebRTC functionality.

Kaufman didn’t think that’s a good idea, and he made no secret of his opinion. “At Microsoft, we have a bit of a more deliberate release cycle,” he said, suggesting that the company wouldn’t support the technology in its consumer products until it’s actually a standard. And, when asked specifically about Google’s support for WebRTC, he replied: “Our companies have a different way of working.”

He likened the current state of WebRTC to WebSocket – a web communications technology that’s supposed to make it easier for HTML5 apps to send messages from servers to browsers, and the other way around. Early versions of WebSocket were widely supported by everyone but Microsoft – but some of the browser makers quickly had to backpedal when a security vulnerability surfaced in 2010.

Microsoft’s take: More flexibilities, no single codec

So what is the difference between Microsoft’s proposal and the technology currently implemented by Google and others? Kaufman likened the way the existing WebRTC proposal works to a black box within the browser: Issues like codecs and the way media is sent over the network are predetermined, with little that app developers can do to optimize for their unique challenges, he explained.

Microsoft, on the other hand, wants to allow more customization and flexibility to implement the technology on legacy devices, Kaufman said. “The history of voice over IP is filled with incompatibility,” he said. Developers of future WebRTC apps should have the power to directly deal with these challenges.

Beyond these issues, there are other philosophical differences between Microsoft and Google, with the question of codecs being a key roadblock. Google and Mozilla want to use VP8, the video codec open sourced by Google in 2010, as the default video codec for WebRTC. Microsoft has been sceptical about VP8 in the past, and Kaufman told me that this stance hasn’t changed. This point was also emphasized in the blog post that announced Microsoft’s proposal Monday. It read, in part:

“Flexibility in its support of popular media formats and codecs as well as openness to future innovation – A successful standard cannot be tied to individual codecs, data formats or scenarios.”

All politics around codecs aside, Kaufman remained optimistic that everyone involved will eventually agree on a common standard. But Microsoft obviously doesn’t want to leave the future of real-time communications up to the Mozillas and Googles of this world, which is why Kaufman urged: “It’s really important for the other participants to get involved in the process.”

Image courtesy of Flickr user  Tsahi Levent-Levi.

11 Responses to “Microsoft commits to WebRTC – just not Google’s version”

  1. Michael Hicks

    So pretty much Microsoft is saying they will continue to hold up the Internets growth cycle due to ignorance and ability to be a team player. I am not surprised and there will come a day when Microsofts greed catches up to them and they find themselves unable to even sell an OS. I love windows but these guy need to really get on the ball especially when they cant even keep up with the big boys and release a browser in less than a 2 year cycle.

  2. WebRTC isn’t a finished spec, neither IETF or W2C have published anything definitive. It’s perfectly reasonable for a participant to contribute alternatives documents for discussion by the groups. Google did this a couple of months back when it drew up the JSEP version which has now replaced the ROAP draft that Cisco wrote earlier. JSEP was an improvement to ROAP and and MS’s new document includes some worthwhile changes.
    The timing is unfortunate, but in general that’s how standards are written.

  3. Airmaxed

    MS is worried that if google has their way, there will be no need for skype, as this technology will cross over and work across all browsers in the future. Also ms, was not built on open source. This web rtc is brilliant, and google is king, ms just keep trying to monitize all things internet, and you will be tailing in the world of open source.

  4. Steve Price

    This is definitely a wrench-in-the-works power play from MS. The WebRTC standard has been in development for TWO YEARS and they chose now, exactly when provisional versions of WebRTC are landing in both Chrome and Firefox, to dump this. Microsoft’s arguments about advocating for more flexibility in the standard is rather spurious, as the established way to improve a standard is to let the standard reach Recommendation status, then add a new layer on top of the existing standard. The whole situation wreaks of Microsoft and and other codec patent-holders trying to preserve their fiefdoms. Not to mention my skepticism about MS moving away from Skype’s walled-garden approach.

  5. Tsahi Levent-Levi

    Besides the issue of slowing down Google with their own implementation, and wanting to squeeze royalties out of H.264 patent licensing, there’s the issue of mindset:

    Google works with internet thinking while Microsoft looks at it with enterprise thinking.

    My own view? If we add more flexibility and options to WebRTC, we’re going to kill it and slow it down instead of improving it:

  6. Martin J. Steinmann

    In the Microsoft usual way they dumped a completely new Microsoft spec into a standardization effort that has been going on for several years. This is arrogant and not collaborative and shows the difference of the Microsoft approach.

    As it relates to codecs of course Microsoft would like to use theirs. Fully proprietary and patent protected, these codecs are the foundation of the Microsoft walled garden. Google’s VP8 is not only open source but came with a free patent grant. Doing so Google transformed a technology into a strategy to change an industry.

    If you want innovation, go with open systems and say no to Microsoft’s walled garden.

    • Brian Burke

      MS isn’t saying we want to lock anyone into anything. They are proposing changes that add flexibility in all of the options as far as codecs etc… They say they want it more open than it is.

      • Except that, as it stands today, the WebRTC standard doesn’t specify that you can *only* use these codecs, they just have a set of codecs that *must* be implemented so that interoperability can be achieved.

        These codecs are VP8 for video (not best choice in quality but you need to have codecs that can be freely licensed in a standard) and g.711 and Opus for audio (Opus has just been approved by the IETF and actually is the best choice for audio).

        So long as Microsoft implement these codecs they can add support for whatever other codecs they want and still interoperate with others.

        Skype already uses VP8 for video and have previously committed to using Opus (their SILK audio codec is part of Opus, which they helped design), so i really don’t see what the problem is.

  7. ʞǝɹɐɯ sɐɯoʇ

    Well, it’s quite clear that voice and video calls codecs are the next standard battle frontier. Microsoft has been aware of that and hoped to grab Nokia, then a strong contender to help it get its way. To take part int the the battle at least, one needs to own a decent hardware manufacturer. Apple produces its own iPhones, Google has Nexues and Redmond has Trojan Horse in Nokia, once number one on the mobile handset market. What else is necessary? A robust cloud-based video-call system. Neither Face Time, nor Skype, however, offer standard, open technology. Since the phone calls must be interoperable, the winner takes it all. Umm, royalties, stupid.