HLS Merge Changes #80
Conversation
…e. Removed extra console calls.
HTML expects buffered end time to be reported relative to the "start" of the video but the netstream reports it relative to the playhead position. In the model, add the buffered value coming from the netstream to the current playtime before returning it to javascript.
Previously, the position of the playhead was being added onto the buffered time both inside the HTTPVideoProvider and the model. Fixed it so that the playhead position is only added once, inside provider.
Instead of requiring an explicit call into the SWF to set the new starting offset for the netstream time, capture it when seekForSeconds is called in HTTPVideoProvider.
In order to distinguish between just running out of buffer and completing the stream when in data generation mode, it's necessary to signal back to the the HTTPVideoProvider. Media sources uses endOfStream to accomplish this so we're exposing that method on the SWF.
|
Could you expand on the title/description? What should be added to the swf changelog when this is pulled in? |
|
I'd like to pull in the latest version of the swf into the 4.4 video.js release. Is there any specific testing that should be done to make sure basic HTTP mp4 playback hasn't changed with this? Also, since PRs require one other person to review, who should review this? Or have both @dmlap and @seniorflexdeveloper reviewed it all already? |
|
@heff: I've reviewed it but still want to do some testing. Will update in here with those results. |
|
@dmlap, this is the change that the release is waiting on, correct? |
|
Per @BCjwhisenant internal comments yesterday, I believe this is verified and good to go. |
|
@seniorflexdeveloper, can you help me by answering this question:
Some kind of summary of features being added would be great. |
|
@heff -
|
|
Awesome, thanks. |
The dist/SWF is the result of the hls-seeking branch merge.