HTML Media Task Force Teleconference -- 10 Dec 2013

<trackbot> Date: 10 December 2013

<paulc> adrian: Are you going to call in?

<paulc> I need you for the MSE discussions.

<scribe> ScribeNick: adrianba

<scribe> Scribe: Adrian Bateman

Roll call, introductions and selection of scribe

paulc: done

MSE status and bugs

paulc: there is an editor's draft and candidate CR draft
... two bugs still open

<paulc> MSE bugs: http://tinyurl.com/6pdnzej

paulc: we'll cover these below

ACTION-57?

<trackbot> ACTION-57 -- Aaron Colwell to File a bug on mse about the file api change -- due 2013-12-10 -- PENDINGREVIEW

<trackbot> http://www.w3.org/html/wg/media/track/actions/57

paulc: this is done

close ACTION-57

<trackbot> Closed ACTION-57.

ACTION-58?

<trackbot> ACTION-58 -- Adrian Bateman to Speak to editors of file api spec about resurrecting "origin of blob" text in spec -- due 2013-12-10 -- OPEN

<trackbot> http://www.w3.org/html/wg/media/track/actions/58

<paulc> Adrian's proposal is in: https://www.w3.org/Bugs/Public/show_bug.cgi?id=23989#c2

adrianba: i updated the bug with a proposal

acolwell: sounds fine with me - i can update the spec

paulc: the place that refers to the text coming in line can be changed

acolwell: we still have a dependency on createObjectURL

paulc: think we can close action-58?

adrianba: yes

close ACTION-58

<trackbot> Closed ACTION-58.

ACTION-59?

<trackbot> ACTION-59 -- Paul Cotton to Coordinate with web apps on the streams api, give them mse requirements and timeline -- due 2013-12-10 -- OPEN

<trackbot> http://www.w3.org/html/wg/media/track/actions/59

paulc: this is pending - doesn't block the next step

adrianba: i'm meeting with one of the editors of stream on friday and can discuss this

action-59 due in 1 week

<trackbot> Set action-59 Coordinate with web apps on the streams api, give them mse requirements and timeline due date to 2013-12-17.

ACTION-60?

<trackbot> ACTION-60 -- Adrian Bateman to Produce a summary of last call comments dispositions -- due 2013-12-10 -- OPEN

<trackbot> http://www.w3.org/html/wg/media/track/actions/60

paulc: any update on this?

adrianba: will send offline

<paulc> it is somewhere in the HTML wiki

Bug 23169 - reconsider the jitter video quality metrics again

paulc: been reopened by Microsoft again

<paulc> Fix: https://www.w3.org/Bugs/Public/show_bug.cgi?id=23169#c16

<paulc> Reopen: https://www.w3.org/Bugs/Public/show_bug.cgi?id=23169#c17

paulc: implementation from F2F discussion was done on dec 2
... jerry re-opened it

jdsmith: we just requested the units be adjusted

acolwell: this was an accident - it was implied to be seconds before and i accidentally changed it to imply microseconds
... i will fix it

<paulc> CfC for MSE CR Working Draft is awaiting ACTION-58, ACTION-60 and Bug 23169

CfC for MSE CR Working Draft is awaiting ACTION-58, ACTION-60 and Bug 23169

paulc: we're waiting on the bug fixes and the ACTION-60 document
... doesn't seem to be anything blocking us from getting candidate CR and running CfC
... need LC disposition of comments for the meeting with the Director
... the right plan is to make the two pending changes for 23169 and 23989 and send a note that the candidate draft is ready to go
... you said when you ran pubrules it complained about implementation experience

acolwell: i think that is a respec nit

paulc: this escalates ACTION-60 - it will be on the critical path
... i will send a note to the team after the meeting with a timeline
... i originally proposed that as soon as we have the candidate draft and run the CfC in the TF and WG at the same time

adrianba: propose we ask for objections from the TF and then run CfC in the WG since the TF is a part of the WG
... if the WG approves then the TF must too

paulc: we could do that if the meeting is okay with that
... any objections?

<paulc> The candidate draft will include the proposed changes to bug 23989 and 23169 discussed at this meeting.

paulc: we will get the CfC started as soon as we get the candidate draft

acolwell: suggestions for date?

paulc: thursday jan 8?

acolwell: and the end date is 3 months after

paulc: yes
... did we agree features at risk?

acolwell: totalFrameDelay
... i will get something published today

<paulc> Paul suggested Thu Jan 8 as the publication date with only frame delay as the CR feature at risk.

EME status and bugs

paulc: agenda item 6 has new bugs
... agenda item 7 has a bug david wanted to discuss
... agenda item 8 documents next steps for bugs
... mostly on the editors
... only thing i want to do here is to give anyone named an opportunity to push back
... propose we spend the next while on item 6, the new bugs

New EME bugs this week

<ddorwin> 23828 and 23897 are related to the addition of keysystem to HTMLSourceElement

Bug 23828 - HTMLSourceElement with |keySystem| attribute should create MediaKeys object for |keySystem|

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23828

https://www.w3.org/Bugs/Public/show_bug.cgi?id=23828

ddorwin: in the original spec we added keysystem attribute
... we found some inconsistencies at TPAC
... this bug is to create a MediaKeys object if there is keysystem

<paulc> 2nd bug related to this topic is https://www.w3.org/Bugs/Public/show_bug.cgi?id=23897

ddorwin: Henri pointed out that the spec only added a script attribute
... so the other bug is to do what we intended and allow content attributes
... this was always the intent but the spec didn't say that correctly
... there is still the question of whether we want to do this
... it is mentioned in the features at risk bug
... this might be the last thing to fix on this before making the decision to keep or not

adrianba: +1

paulc: are you saying wait until just before LC?

ddorwin: yes, this will be at the bottom of the list to fix

<paulc> Both bugs 23897 and 23828.

Bug 23866 - Remove section 4.3. Addition to Media Element Load Algorithm

https://www.w3.org/Bugs/Public/show_bug.cgi?id=23866

ddorwin: a tracking bug to clean up spec text
... this is about separating the MediaKeys from the media element
... you should be able to change the source or reload and the MediaKeys remains attached

Bug 23955 - Add optional MediaKeys.loadSession(DOMString sessionId) method

https://www.w3.org/Bugs/Public/show_bug.cgi?id=23955

ddorwin: proposal to make a peer to createSession to take a session ID and load the session
... lots of discussions about this - for example key release
... this is a proposal to formalise this and define how it looks including same origin rules, etc.

<markw> yep

ddorwin: realising this isn't always required it is optional

paulc: i note that markw says this could allow key release bug to be closed
... any discussion on this?

Bug 24025 - Add optional configuration parameter to MediaKeys constructor

https://www.w3.org/Bugs/Public/show_bug.cgi?id=24025

ddorwin: looking at other use cases (e.g. offline specific licenses)
... we don't have a way to pass those - it could be done in the keymessage/update path it would require extra round trips
... propose to add a configuration parameter to the CDM
... mark suggested a dictionary
... all of this would be optional - not sure we can define all the parameters
... Henri replied this morning - i will reply later

paulc: mark also responded

markw: just a question about the dictionary
... i think this is to say this is a javascript object and doesn't prevent additional values
... implementations are free to ignore undefined members

Bug 24026 - Document that URLs (destinationURL) should/may be validated and/or normalized

https://www.w3.org/Bugs/Public/show_bug.cgi?id=24026

ddorwin: URLs can come from CDM or potentially from media data
... in some implementations they get normalised
... we at least need to say they will be normalised or validated
... there is a question of if we say they should be
... i think we might have discussed this before and one issue was how to define normalise
... we could include a warning that it might be different across browsers
... i know there is a URL spec - perhaps it will help

paulc: does anyone know the status of this in W3C and IETF?
... think no work currently at W3C

ddorwin: least controversial is they may be validated, next is they should be validated, and finally is switching to some kind of URL type which seems further down the road

pal: what is the exact definition of validated?

ddorwin: i would assume either invalid url components may be dropped or you might lose the entire url

pal: so it looks like the question is what is the defining document for a valid URL
... i guess changing a URL out from under a piece of code that might expect something specific sounds impolite

adrianba: i think it is okay to change the url given its purpose
... but unless there is a concrete definition of validate or normalize i think the distinction between must and should isn't useful

pal: the URL may have come from the media data, right?

ddorwin: yes

pal: if this is going to be transformed it would be good to know how
... this is going to be in files - if it gets transformed differently in different implementations
... we should know exactly what happens

ddorwin: this is about nits in URLs - if you create a valid URL that is valid in most browsers then you won't have a problem
... it's intended to be used in XHR

pal: making sure that the file contains a valid URL is fine but i think we need to be very precise

paulc: anne's spec is supposed to document what browsers do
... one last item i want to process

Other EME bugs for discussion

Bug 23733 - Consider prohibiting support of active content by CDMs

https://www.w3.org/Bugs/Public/show_bug.cgi?id=23733

paulc: wondering what we want to do about this
... don't think we have time for the discussion today

markw: the bug was to consider this topic - it was proposed originally in the security section by glenn i think
... there was a lot of pushback at tpac, especially the definition of active
... if there was a way to prohibit executable content - i think there is a class of content nobody wants - then we could define that
... we could frame this as a question to CDM vendors
... in the absence of any consensus then we can drop this

paulc: i'm trying to understand what change you are proposing

markw: i'm not sure what is in there next

paulc: why don't you take the f2f and make a proposal?

markw: not sure i can do that

paulc: doesn't that mean won't fix?

adrianba: we could close NEEDSINFO but i think mark was proposing to ask for a discussion first

Other EME bugs awaiting action by Editors

paulc: lots of action items for editors

markw: i think 17199 is dependent on loadSession bug

Chair and Scribe for next meeting

acolwell: i pushed new MSE and candidate CR while you were talking about EME
... we should be good to go

paulc: i will look at CfC today - ACTION-60 is on the critical path

Adjournment

paulc: think we can adjourn - will have EME next week
... david thanks for your suggestions for the agenda
... adjourned

- DRAFT -

HTML Media Task Force Teleconference

10 Dec 2013

Attendees

Contents