Chromium Code Reviews| Index: remoting/proto/video.proto |
| diff --git a/remoting/proto/video.proto b/remoting/proto/video.proto |
| index 0ee29c9b8922d65c30dfb653fa32db538104eb79..211b0231ddbeae5c1b2b460f6244cded5fcbf6aa 100644 |
| --- a/remoting/proto/video.proto |
| +++ b/remoting/proto/video.proto |
| @@ -66,4 +66,16 @@ message VideoPacket { |
| // Optional frame timestamp. Used in tests to estimate frame latency. |
| optional int64 timestamp = 12; |
| + |
| + // Frame identifier used to match VideoFrame and VideoAck. |
| + optional int32 frame_id = 13; |
| +} |
| + |
| +// VideoAck acknowledges that the frame in the VideoPacket with the same |
| +// frame_id has been rendered. VideoAck messages must be sent only for frames |
| +// that have frame_id field set. They must be sent the same order in which |
| +// the corresponding VideoPackets were received. |
|
Wez
2015/02/11 02:22:56
Note that if we move to an unordered transport the
Sergey Ulanov
2015/02/17 19:37:06
It will be easy to remove this requirement for uno
Wez
2015/02/21 03:12:02
Acknowledged.
|
| +message VideoAck { |
|
Wez
2015/02/11 02:22:56
nit: You could almost call this VideoFrameFeedback
Sergey Ulanov
2015/02/17 19:37:06
I think VideoAck is fine for now - it can be renam
Wez
2015/02/21 03:12:02
Acknowledged.
|
| + // Frame ID. |
| + optional int32 frame_id = 1; |
| } |