Introduce struct MediaConfig, with construction-time settings.

talk/app/webrtc/mediacontroller.h

 /*
  * libjingle
  * Copyright 2015 Google Inc.
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions are met:
  *
  *  1. Redistributions of source code must retain the above copyright notice,
  *     this list of conditions and the following disclaimer.
  *  2. Redistributions in binary form must reproduce the above copyright notice,
@@ skipping 20 matching lines @@
 #include "webrtc/base/thread.h"
 
 namespace cricket {
 class ChannelManager;
 }  // namespace cricket
 
 namespace webrtc {
 class Call;
 class VoiceEngine;
 
+// Construction-time settings, passedon when creating MediaChannels.

[pthatcher1, 2016/02/05 17:11:30]

passedon => passed on

[nisse-webrtc, 2016/02/08 09:02:32]

Done.

+struct MediaConfig {
+  // Set DSCP value for packet sent from media channel. This flag

[pthatcher1, 2016/02/05 17:11:30]

DSCP value for packet => DSCP on packets 
from media channel => (deleted)

[nisse-webrtc, 2016/02/08 09:02:32]

Done.

+  // comes from the PeerConnection constraint 'googDscp'.
+  bool enable_dscp = false;
+
+  // Video-specific config
+
+  // Enable WebRTC Cpu Overuse Detection. This flag comes from the

[pthatcher1, 2016/02/05 17:11:30]

Cpu => CPU

[nisse-webrtc, 2016/02/08 09:02:32]

Done.

+  // PeerConnection constraint 'googCpuOveruseDetection' and is
+  // checked in WebRtcVideoChannel2::OnLoadUpdate, where it's passed
+  // to VideoCapturer::video_adapter()->OnCpuResolutionRequest.
+  bool cpu_overuse_detection = true;

[pthatcher1, 2016/02/05 17:11:30]

Please provide a blank line here

[pthatcher1, 2016/02/05 17:11:30]

To be consistent, shouldn't this be enable_cpu_overuse_detection?

[nisse-webrtc, 2016/02/08 09:02:32]

Makes sense. (I just moved it over from VideoOptions without edits to name or
comments).

+  // Set to true if the renderer has an algorithm of frame selection.
+  // If the value is true, then WebRTC will hand over a frame as soon as
+  // possible without delay, and rendering smoothness is completely the duty
+  // of the renderer;
+  // If the value is false, then WebRTC is responsible to delay frame release
+  // in order to increase rendering smoothness.
+  //
+  // This flag comes from PeerConnection's RtcConfiguration, but is
+  // currently only set by the command line flag
+  // 'disable-rtc-smoothness-algorithm'.
+  // WebRtcVideoChannel2::AddRecvStream copies it to the created
+  // WebRtcVideoReceiveStream, where it is returned by the
+  // SmoothsRenderedFrames method. This method is used by the
+  // VideoReceiveStream, where the value is passed on to the
+  // IncomingVideoStream constructor.
+  bool disable_prerenderer_smoothing = false;
+};
+
 // The MediaController currently owns shared state between media channels, but
 // in the future will create and own RtpSenders and RtpReceivers.
 class MediaControllerInterface {
  public:
   static MediaControllerInterface* Create(
       rtc::Thread* worker_thread,
+      const MediaConfig& config,

[pthatcher1, 2016/02/05 17:11:30]

I think it should be the first argument.

[nisse-webrtc, 2016/02/08 09:02:32]

Done.

       cricket::ChannelManager* channel_manager);
 
   virtual ~MediaControllerInterface() {}
   virtual webrtc::Call* call_w() = 0;
   virtual cricket::ChannelManager* channel_manager() const = 0;
+  // TODO(nisse): Reference or pointer?

[pthatcher1, 2016/02/05 17:11:30]

Reference (or copy; it's so small it doesn't matter)

[nisse-webrtc, 2016/02/08 09:02:32]

Done.

+  virtual const MediaConfig* config() const = 0;
 };
 }  // namespace webrtc
 
 #endif  // TALK_APP_WEBRTC_MEDIACONTROLLER_H_