ContainerParser currently assumes that each append starts on a random access point (e.g. init segment or cluster) and contains a complete chunk. That's not something that the spec requires so ContainerParser must be extended to handle non-aligned appends. This is important right now because YouTube use this to append ~64kB chunks as a way of controlling the streaming rate of XmlHttpRequests.
5 years ago
Same as before, but with cruft excluded.
Comment on attachment 8482186 [details] [diff] [review] Handle unaligned appends by retaining WebM parser state across calls. r=cajbir Still has cruft.
Attachment #8482188 - Flags: review?(cajbir.bugzilla) → review+
(In reply to Matthew Gregan [:kinetik] from comment #6) > Created attachment 8482449 [details] [diff] [review] > Only start new decoder on media segment boundary. What does this patch apply on top of? I get conflicts with SourceBuffer.cpp in ParseStartAndEndTimestamps.
Attachment #8482449 - Flags: review?(cajbir.bugzilla) → review+
Status: ASSIGNED → RESOLVED
Closed: 5 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla35
You need to log in before you can comment on or make changes to this bug.