[webkit-reviews] review granted: [Bug 213369] Add logging to WebRTC video pipeline to check for frame rate stability : [Attachment 402544] Patch

[bugzilla-daemon at webkit.org]

Eric Carlson <eric.carlson at apple.com> has granted youenn fablet
<youennf at gmail.com>'s request for review:
Bug 213369: Add logging to WebRTC video pipeline to check for frame rate
stability
https://bugs.webkit.org/show_bug.cgi?id=213369

Attachment 402544: Patch

https://bugs.webkit.org/attachment.cgi?id=402544&action=review




--- Comment #10 from Eric Carlson <eric.carlson at apple.com> ---
Comment on attachment 402544
  --> https://bugs.webkit.org/attachment.cgi?id=402544
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=402544&action=review

>
Source/ThirdParty/libwebrtc/Source/webrtc/video/rtp_video_stream_receiver.cc:74
5
> +    static const unsigned MaxFrameDelayCount = 3;

Why is this const and the constants below are constexpr?

>
Source/ThirdParty/libwebrtc/Source/webrtc/video/rtp_video_stream_receiver.cc:74
6
> +    static constexpr unsigned TimeStampStorageDurationInMs = 2000;

Maybe TimeStampQueueDurationInMs?

>
Source/ThirdParty/libwebrtc/Source/webrtc/video/rtp_video_stream_receiver.cc:76
4
> +    auto interval = frameTime - observedFrameTimeStamps_.front();

Maybe "queueDuration" instead of "interval"?

> Source/WebCore/platform/FrameRateMonitor.h:44
> +    double frameRate() const { return m_observedFrameRate; }

Should this be "observedFrameRate"?

> Source/WebCore/platform/FrameRateMonitor.h:50
> +    static constexpr Seconds MinimumAverageDuration = 1_s;
> +    static constexpr Seconds TimeStampStorageDuration = 2_s;
> +    static constexpr unsigned MaxFrameDelayCount = 3;

Is there a benefit to declaring these here rather than where they are used as
you do with the constants in RtpVideoStreamReceiver?