Reflections on the WebRTC Conference

By: Eric Hernaez on December 3rd, 2012

This past week, Chris Aaker (NetSapiens’ Director of Engineering) and I headed to San Francisco to attend the WebRTC Conference and Expo.  

In case there was any doubt, the event made clear that WebRTC is a disruptive technology that will transform the way people use the Internet in years to come.   The sheer  density of industry thought leaders, and the shared understanding that we we’re in the midst of something big, created an energy that I haven’t experienced since the  early days of the VoIP industry.  Kudos to Dave Rodriguez and the rest of the TMC team for putting together a great event.

Here are some observations from my three days in South San Francisco (in no particular order):

    • WebRTC is a reality with full support for audio/video peer sessions in Chrome 23 (released on Nov 7, 2012), and availability in Mozilla Firefox 19 promised  for January 2013.  When that happens 58% of browser users will have access to it.
    • There was no shortage of grand predictions.  Google’s Hugh Finnan, director of product management for Chrome, reflects the sentiment of many others when he states that  WebRTC  “has the potential to be as important to the Web as HTML was in the beginning.”
    • There are disagreements; particularly about whether  codecs should be mandatory.  For video codecs, support is divided between H.264 and VP8.   Microsoft and others are in favor of H.264 while Google and Mozilla prefer VP8 (because it is royalty free).  Microsoft people are very active in the standards committees and have proposed a competing standard, RTCWeb, which would remove VP8 as the mandatory video codec, among other things.  See here for a good discussion about the differences between the competing proposals.
    • Interestingly, hardware vendors who implement video encoding/decoding in chipsets  find themselves in the Microsoft camp simply because there are few hardware implementations of VP8 to date.  Although the VP8 codec is licensed for free as part of Google’s WebM project, only a handful of silicon makers have productized it.
    • Despite the rift with Microsoft, the consensus among the Google developers and product managers is that Microsoft must be interoperable with the Firefox and Chrome implementations because it needs to support Skype users across all browsers.
    • Everybody agrees that WebRTC support for mobile devices is imperative and Google expects to have it Android within 2013.   Supporting WebRTC in a mobile browser is challenging because of the battery and CPU constraints inherent to mobile devices.
    • Unlike Microsoft, Apple is not active in the WebRTC standards committees, and nobody knows what Apple is planning to do.  That said, such behavior is the norm for Apple (when was the last time that Apple pre-announced anything?), and several panelists cautioned that Apple’s lack of involvement should not be considered as evidence in either direction regarding support for WebRTC in Safari and Mobile Safari (Webkit).
    • The current approach to Apple’s unknown intentions for WebRTC support in Mobile Safari is to talk about a C++ WebRTC library that can be compiled into a native iOS app.  Indeed, Ericsson has released Bowser, a mobile browser that currently works on iPhone and iPad devices, and is available for download from the App Store.  For more info, see the Ericsson Labs website.
    • WebRTC effectively turns any browser into a audio/video terminal, and therefore it will replace many desktop phones – especially in contact centers.  The real value of WebRTC is not in the ability to replace a desktop phone, however, or even in the capability to add video to a communications session.  The real value lies in the ability to give contact center employees context for a communication session.  For example, if I click the “Call” button while shopping for a power tool at Sears.com, my session can be routed directly to someone with knowledge of power tools.  This facet alone will revolutionize the way that commerce is done online.
    • The general consensus among panelists was that WebRTC will accelerate the ongoing shift of power from enterprise telecom departments to enterprise IT departments.  Several attendees questioned whether WebRTC frameworks might completely replace telephony in customer contact environments.  The panelists fielding these questions (from Mitel, Avaya and Seimens) did not see that happening.  I tend to agree.  Until the web completely replaces the PSTN (sometime in the year 2165), telephony infrastructure will be closely intertwined with the web.
    • There were dozens of demos throughout the conference.  Almost all of them used the same 20+ lines of code as the initial demo video calling app released  earlier this year.   In my estimation, more than 2/3 of the demos failed to the point where it became a running joke among the demo-ers and demo-ees alike.   The award for most graceful demo fail goes to Jose De Castro of Voxeo Labs.
    • With the sole exception of Ericsson’s Martin Körling, very little was said about the potential for WebRTC in web TV.  However, the possibility to create apps that leverage interactive TV right through the browser is very exciting.
    • Despite the availability of early SIP-WebRTC gateways  – it is supported in upcoming versions of  Asterisk (11) and Kamailio (3.4) – there are a lot of issues to be worked out before WebRTC “apps” can be considered truly interoperable with the SIP world.  Nonetheless, every SIP platform vendor at the event discussed products that would leverage WebRTC.  Although there were several demos showing calls gatewayed between the PSTN and a browser, nobody has a product ready for GA today.

In case you are not familiar with WebRTC, this video from Cullen Jennings of Cisco does a good job explaining what it is.