Closed avelad closed 2 years ago
I've pinned this to the top of the issues list and renamed it. Please, everyone, let us know your top priorities for v2.7+.
For us, the most important Functionality, we cant wait to use is any kind of Preload API. Far above anything else.
I'd like to see a solution to #2179. Then we could remove our workaround.
Not sure if that will be covered by your "Codec-switching" task
Regards, Rik.
For us, the most important functionality is the CMAF low latency. Regards
Please note that although our plans & timelines have not changed, we have decided to rename v2.6 to v3.0, and therefore v2.7 planning has become v3.1 planning.
For us the most important values related to video playback are performance and stability. So if in future releases could be more effort spent on that we would welcome that. To highlight some ideas (not in order).
100% Low-latency live (LHLS following Apple specs, equivalent for DASH) for us
@joeyparrish Are there already selected priorities?
Thank you, everyone, for the feedback!
At this time, we are prioritizing LL-HLS, followed by MediaCapabilities + codec-switching. We have a design for LL-HLS from @michellezhuogg, who will also work on the implementation.
Will low latency be supported in DASH?
At a minimum, we will add support for the availabilityTimeOffset
attribute for low-latency DASH. We are debating the value of also using ReadableStream
to pipeline data from the network into MediaSource for LL-DASH. This would require major changes to NetworkingEngine
and the scheme plugin API, so we want to make sure there is a measurable latency improvement before we commit to a breaking change like that.
There is a DASH-IF presentation where this topic is discussed. Really if chunks are not passed directly to the media source, there is really no low latency, because chunks would not be used at all.
https://dvb.org/wp-content/uploads/2020/03/Latest-on-DASH-low-latency.pdf
According to the previous image, which case is the one you want to implement?
It would be great to have:
For us would be the ability to play media playlists #2300
@avelad, let's discuss details of DASH low-latency on #1525.
It would be awesome to have:
1) Codec-switching 2) Be clever and choose a stream resolution adapted to the player size (to save bandwidth if the player is tiny) 3) Bundle size reduction
@avelad @joeyparrish would you be interested to maybe link to issues opened per item in the roadmap and we can then collect thumbs up on each to see community traction? I would really appreciate this and enjoy if we have this.
Would love to see
Would also love to see Preload API, with support for seamless playback between two dash manifests.
Hi everyone,
At this point, we are nearly done with v3.1, which will include LL-HLS and LL-DASH, plus our own CEA 608/708 parser to replace mux.js. I will rename this issue to continue to collect feedback on priorities for v3.2, for which work has not yet started.
Codec switching/preference would be great for 3.2 - https://github.com/google/shaka-player/issues/2179
Smooth streaming support is a priority to be competitor with other players like dash.js #703
Given that the development of version 2.6 is coming to an end, I would like to be able to open the discussion of the things that are most priority for the integrators of the ShakaPlayer.
To limit the list each member should propose 3 things and in order of priority.
The items to pick should be the entries that appears in https://github.com/google/shaka-player/blob/master/roadmap.md
For me, the priority is: 1) Thumbnail tracks 2) Low-latency live (LHLS, equivalent for DASH) 3) Codec-switching