For the result of `VIDEO_PLAY_TIME_MS`, it seems reasonable to me. The previous collective way would sometime result in reporting same accumulated result multiple times. Eg. a video has already been played 10s and it triggers the point of reportting telemetry (report 10s). Then it keeps playing to 20s and trigger another point (report 20s). So we actually report 30s for the media that only plays for 20s. The new triggering point of reporting telemetry will happen whenever the media decoder stops decoding. Eg. pause, seek, resolution change, or media gets destroyed and it won't report the same value again. Eg. play a video that is 30 minutes long, but pause on the place at 5 minutes for a break (report 5 minutes), and then go back to watch the rest of 25 minutes (report 25 minutes) So I expect to see (1) more reported sample count and (2) each sample count has less value because we won't accumulate data. You can see both situations happened on the telemetry report, so I think it's no problem. --- For the result of `VIDEO_WIDEVINE_PLAY_TIME_MS`, the result looks a little bit suspicous. I tested on [this video](https://shaka-player-demo.appspot.com/demo/#audiolang=en-US;textlang=en-US;uilang=en-US;asset=https://storage.googleapis.com/shaka-demo-assets/angel-one-widevine-hls/hls.m3u8;panel=ALL_CONTENT;build=uncompiled) and the result looks the same as the real time that I spent on watching the video. Because the result looks too low, I wonder if we [can't get tFor the result of `VIDEO_WIDEVINE_PLAY_TIME_MS`, the result looks a little bit suspicous. I tested on [this video](https://shaka-player-demo.appspot.com/demo/#audiolang=en-US;textlang=en-US;uilang=en-US;asset=https://storage.googleapis.com/shaka-demo-assets/angel-one-widevine-hls/hls.m3u8;panel=ALL_CONTENT;build=uncompiled) and the result looks the same as the real time that I spent on watching the video. Because the result looks too low, I wonder if we [can't get the key system](https://searchfox.org/mozilla-central/rev/a9851cca236e03d088d7528dd27445080475a68e/dom/media/utils/TelemetryProbesReporter.cpp#189-200) at the time we report the result to the Telemetry, which makes us not be able to report the CDM play time. he key system](https://searchfox.org/mozilla-central/rev/a9851cca236e03d088d7528dd27445080475a68e/dom/media/utils/TelemetryProbesReporter.cpp#189-200) at the time we report the result to the Telemetry, which makes us not be able to report the CDM play time. Bryce, any thought?
Bug 1685399 Comment 28 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
For the result of `VIDEO_PLAY_TIME_MS`, it seems reasonable to me. The previous collective way would sometime result in reporting same accumulated result multiple times. Eg. a video has already been played 10s and it triggers the point of reportting telemetry (report 10s). Then it keeps playing to 20s and trigger another point (report 20s). So we actually report 30s for the media that only plays for 20s. The new triggering point of reporting telemetry will happen whenever the media decoder stops decoding. Eg. pause, seek, resolution change, or media gets destroyed and it won't report the same value again. Eg. play a video that is 30 minutes long, but pause on the place at 5 minutes for a break (report 5 minutes), and then go back to watch the rest of 25 minutes (report 25 minutes) So I expect to see (1) more reported sample count and (2) each sample count has less value because we won't accumulate data. You can see both situations happened on the telemetry report, so I think it's no problem. --- For the result of `VIDEO_WIDEVINE_PLAY_TIME_MS`, the result looks a little bit suspicous. I tested on [this video](https://shaka-player-demo.appspot.com/demo/#audiolang=en-US;textlang=en-US;uilang=en-US;asset=https://storage.googleapis.com/shaka-demo-assets/angel-one-widevine-hls/hls.m3u8;panel=ALL_CONTENT;build=uncompiled) and the result looks the same as the real time that I spent on watching the video. Because the result looks too low, I wonder if we [can't get tFor the result of `VIDEO_WIDEVINE_PLAY_TIME_MS`, the result looks a little bit suspicous. I tested on [this video](https://shaka-player-demo.appspot.com/demo/#audiolang=en-US;textlang=en-US;uilang=en-US;asset=https://storage.googleapis.com/shaka-demo-assets/angel-one-widevine-hls/hls.m3u8;panel=ALL_CONTENT;build=uncompiled) and the result looks the same as the real time that I spent on watching the video. Because the result looks too low, I wonder if we [can't get the key system](https://searchfox.org/mozilla-central/rev/a9851cca236e03d088d7528dd27445080475a68e/dom/media/utils/TelemetryProbesReporter.cpp#189-200) at the time we report the result to the Telemetry, which makes us not be able to report the CDM play time. Bryce, any thought?