<trackbot> Date: 11 June 2013

<scribe> scribenick: adrianba

<scribe> scribe: Adrian Bateman

<acolwell> http://lists.w3.org/Archives/Public/public-html-media/2013Jun/0012.html

Roll call, introductions and selection of scribe

acolwell: done

Review of action items and issues

acolwell: none

MSE status and bugs

acolwell: since the last call i published 2 spec updates
... on jun 1 and jun 5
... tried to tackle as many of the pre-LC bugs as i could address
... please take a look
... still 11 bugs outstanding to discuss

<paulc> Sorry I was late. Computer problems.

acolwell: questions?

<scribe> Chair: Paul Cotton

<paulc> Link to open bugs: https://www.w3.org/Bugs/Public/buglist.cgi?query_format=advanced&product=HTML%20WG&component=Media%20Source%20Extensions&bug_status=NEW&bug_status=ASSIGNED&bug_status=REOPENED&list_id=15288

-> http://tinyurl.com/6pdnzej

Open bugs

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=22148

Bug 22148 https://www.w3.org/Bugs/Public/show_bug.cgi?id=22148

Request that we reconsider adding jitter to video quality metrics

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=22148#c3

paulc: Most recent comment is from Jerry
... this has a specific proposal right in it

acolwell: this got posted yesterday so i haven't had a bunch of time to think about it
... what is the application going to do with this information
... if you're starting to see large delays then you should see the dropped count increase

<paulc> Jerry is out this week, so Adrian is responding.

<paulc> This gives another indication of the quality of the presentation and the app will want to fall back to a lower quality media file.

acolwell: the other question is could this expose user agent specific behaviour?
... if different UAs use different delays for when they consider to drop a frame, could that be problematic
... is this something people are worried about?

markw: from what we understand, UAs could take different approaches before dropping so this metric allows you to detect that there are problems with playback before you get to dropped frames

paulc: mark, do you support adding this feature request?

markw: yes, because you need this in addition to dropped frames to allow consistent behaviour

paulc: shall we move on to another and allow more time to think about this?

Cyril: why is this on the video quality element and not on the media source object?

acolwell: the reason this was added to the video quality object is because they are not MSE specific
... considered as an initial proposal to HTML5 and applies whether MSE is used or not

<paulc> Bu 22138: https://www.w3.org/Buggs/Public/show_bug.cgi?id=22138

acolwell: waiting for the person to respond

adrianba: recommend resolving NEEDSINFO until we get the information needed to make the bug actionable

paulc: i will add that note to the bug

Bug 22137: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22137

acolwell: this is about changing the number of tracks during playback
... this has come back a number of times
... i tried to add a proposal to not have it become too complicated

<paulc> June 11 reply: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22137#c4

acolwell: suggestion is to allow it to be accepted but not guaranteed to be played
... looks like the filer is okay with this
... but want the group to decide if this is something we want to do this version or not
... haven't discussed as a group

<paulc> We request the W3C relax the restrictions in section 11 that “apply to all initialization segments in a byte stream” – particularly requirements #1 (“The number and type of tracks must be consistent”) and #3 (“Track IDs must be the same across initialization segments if the segment describes multiple tracks of a single type”) – at least for audio tracks.

paulc: this is the key questions

<paulc> Current spec is in section 12.

Cyril: you talk about a default flag in the track - what does this mean?

acolwell: in mpeg4 there is a default track flag, i think, not sure which box it is - tracks can be marked as the default for playback
... most formats supporting multiple tracks have a way to indicated which should be played by default

Cyril: you may have multiple tracks enabled

acolwell: if you don't have a default then you pick the first in the init segment

BobLund: not sure i understand the solution
... UA does not need to support multiple audio and video tracks?

acolwell: no, in a single sourcebuffer, if there is one audio and one video track
... then you add another segment with 2 audio and 2 video tracks
... then the UA isn't required to expose the new tracks
... different to the current track, which would reject playback all together
... suggesting that don't reject it but don't guarantee playback

BobLund: so if the first init segment exposes multiple tracks then all have to be available?

acolwell: yes
... we'd need to decide what happens if the number reduces
... the text of the spec was written assuming the number doesn't change

Cyril: you want to bound the number of tracks by the number in the first init segment?

acolwell: my concern is that if things change during playback it is hard to ensure resources are available during playback
... so during the first init this is the time the media engine can reserve the appropriate resources
... and it is optional to add new resources later
... even if there are multiple tracks in the first init segment then the UA isn't necessarily required to support them all

BobLund: i think the HTML spec in the media section defines what UAs should do with multiple tracks
... and i am happy with your proposed solution

<paulc> Q from correspondent: Does an HTML5 UA have a concept of user preferences for audio language or accessibility?

<paulc> Re above question: bob said he did not know if such guidance existed.

<paulc> Adrian: Microsoft needs to check on acolwell's proposal.

<paulc> Adrian: Just need more time to review.

Bug 22136: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22136

<paulc> See https://www.w3.org/Bugs/Public/show_bug.cgi?id=22136#c8

acolwell: i think they accept the proposal so i just need to make the change

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=22135

<paulc> Ade's response: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22135#c4

acolwell: this is a request for track switching that is more general than mse
... i don't think we should change the spec at this point

<paulc> Re solution to 22136 is the support mandatory?

adrianba: on 22136, is this support mandatory?

acolwell: i was trying to specify a mandatory base line so that you don't get a decode error depending on if support is there
... i think from chrome's pov i think what i described worked

<paulc> To be clear we are talking about 22136 here.

<paulc> We will hold off implementing the fix for 22136 and Adrian will respond on the bug with his position.

<paulc> Back 22135:

paulc: back on 22135

<paulc> Ade's response: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22135#c4

paulc: does anybody else want to speak to this?
... done

Bug 22134: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22134

acolwell: need to think about this some more
... spec isn't clear on when content is rejected

paulc: that one is in your court then

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=22117

paulc: bug says please provide text and suggested location

acolwell: still not quite sure what cyril is looking for

Cyril: i looked at all the must statements and most are related to UA
... but some others are related to media segments
... media segments must have this or that
... which seemed like normative statements for generation of content

acolwell: i don't understand the difference between what a UA must do to play it vs. to create it
... the spec outlines what happens in the UA

paulc: we could reword so it says the UA must do something with the media segment

Cyril: worried about people suggesting creating segments that are not conforming

acolwell: they would be segments that MSE UAs can't play

Cyril: was looking at this for writing conformance tests

adrianba: happy to edit the changes into the spec if someone writes them up

Cyril: i can do that

paulc: Cyril will identify the new text - change A to B

Cyril: shall i add them to the bug?

paulc: yes

22112: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22112

paulc: editorial, assigned to adrianba

22110: https://www.w3.org/Bugs/Public/show_bug.cgi?id=22110

acolwell: also editorial

adrianba: wasn't sure what text to put there

acolwell: if the JS layer wants to indicate a decode error this is how
... don't want to enumerate all the reasons

paulc: cyril, can you give us an example

Cyril: think the text that an app can signal a decode error but not sure when this would be used at all

acolwell: if the app does any parsing and doesn't get what it expects it would use this

Cyril: why wouldn't it get another segment

acolwell: MSE replaces the network layer and that layer is able to provide decode and network errors
... and this provides that mechanism in MSE

adrianba: assign to me and i'll update the text

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=22109

22109 - https://www.w3.org/Bugs/Public/show_bug.cgi?id=22109

acolwell: this is editorial, just need to figure out what to say

adrianba: would like to suggest that when someone files an editorial bug suggesting to rename something when please use propose a new name

acolwell: i will think about this

Bug 21431 https://www.w3.org/Bugs/Public/show_bug.cgi?id=21431

paulc: did the changes but not marked as resolved

<paulc> See https://www.w3.org/Bugs/Public/show_bug.cgi?id=21431#c14

acolwell: glenn's comment was here are things that need to be done but there will be more
... so i left it open for glenn to add any more

paulc: recommend sending a personal note to glenn
... not clear if we will have resolved all bugs in 2 weeks
... when we do that we will do a CfC inside the TF to go to LC
... you will have seen my regrets for next tuesday - i come back the following monday so either you get a late agenda, someone else does it, or i do it today as a proforma

adrianba: i'm fine with the day before

acolwell: i'm fine with that too

Adjournment

paulc: adjourned

This is scribe.perl Revision: 1.138  of Date: 2013-04-25 13:59:11  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/???/Cyril/
Succeeded: s/Bu/Bug/
Found ScribeNick: adrianba
Found Scribe: Adrian Bateman
Default Present: davide, Michael_Thornburgh, joesteele, pladd, Cyril, johnsim, Aaron_Colwell, markw, adrianba, BobLund, ddorwin, paulc
Present: davide Michael_Thornburgh joesteele pladd Cyril johnsim Aaron_Colwell markw adrianba BobLund ddorwin paulc
Agenda: http://lists.w3.org/Archives/Public/public-html-media/2013Jun/0012.html
Found Date: 11 Jun 2013
Guessing minutes URL: http://www.w3.org/2013/06/11-html-media-minutes.html
People with action items: