Chrome

Show Contents

• FAQ about the current implementation
  • What’s the status of data channels?
  • What API can I use for screensharing?
  • Native WebRTC Logging in Chrome

Both getUserMedia and RTCPeerConnection are implemented and shipping in Chrome for Windows, Linux, Mac and Android.

FAQ about the current implementation

What’s the status of data channels?

The standards-compliant SCTP-based DataChannel implementation has been released in Chrome 32. You do not need to specify any constraint in order to use the SCTP-based DataChannels in Chrome.

A few specific notes on data channels:

• In the standalone WebRTC build, DTLS must be enabled by either the “DtlsSrtpKeyAgreement:true” constraint or providing your own implementation of DTLSIdentityServiceInterface to PeerConnectionFactoryInterface::CreatePeerConnection.

• You should not try to send more than 16KB at a time via the DataChannel.send() API. This limitation is temporary and will be removed in a future update to the SCTP protocol (EOR + ndata). Until then, you can work around this by breaking your data into < 16KB chunks and sending each chunk individually.

• Chrome also supports unreliable RTP-based data channels, and they can be activated by a { RtpDataChannels: true } constraint to the PeerConnection constructor. However, they should be considered deprecated and NOT used in new code. Chrome will remove support for RTP-based data channels at some point in the future.

What API can I use for screensharing?

chrome.desktopCapture.chooseDesktopMedia is the API that you want to use, and is available starting with Chrome 34. This API is currently only available to Chrome apps/extensions, but a web page can use postMessage to communicate with such an extension. For more information on this API, see this discuss-webrtc thread.

Native WebRTC Logging in Chrome

Turning on logging in Chrome in order to debug WebRTC related issues can be very useful. It is furthermore possible to filter the log output to libjingle and webrtc in order to reduce the noise.

When launching Chrome from the command line, here’s an example for how to turn on logging and filter it to WebRTC and libjingle:

chrome.exe --enable-logging --vmodule=*/webrtc/*=2,*/libjingle/*=2,*=-2 --no-sandbox

This example is from Windows, so please replace chrome.exe with the path to the main Chrome binary on the platform you will be testing on (e.g. on Mac this would typically be Google\ Chrome.app/Contents/MacOS/Google\ Chrome). Note also that the --no-sandbox parameter is currently only needed on Windows and since it turns off Chrome’s security sandbox, please use only for testing trusted code.

The logging level you can assign is the verbosity level with higher values being more verbose. More details here.

In order to actually see the logs, it’s fairly simple on Mac and Linux, just watch the output in the terminal, but on Windows, I recommend Sawbuck. Sawbuck has the nice option of being able to get stack traces for each trace, if you have the symbols (and performance is still pretty good). Configure the Chrome provider to “All” and uncheck “Text Only” in the “Enable Mask” column, and hit Ctrl+E to start capturing.