{minutes} HTML WG media telecon 2013-10-15 - EME status and bug discussion

http://www.w3.org/2013/10/15-html-media-minutes.html

Joe Steele
steele@adobe.com




HTML Media Task Force Teleconference

15 Oct 2013

Agenda

See also: IRC log

Attendees

Present
paulc, Niels_Thorwirth, pladd, joesteele, davide, +1.425.936.aaaa, ddorwin, glenn, pal, adrianba, [Microsoft], BobLund, +1.425.605.aabb
Regrets
Chair
Paul Cotton
Scribe
joesteele
Contents

Topics
Agenda
previous minutes
EME status and bugs
Davids Open Bug Summary
Bugs on the goal section
Bugs from last meeting
Davids Open Bug Summary (part 2)
Summary of Action Items
<trackbot> Date: 15 October 2013
<paulc> Agenda: http://lists.w3.org/Archives/Public/public-html-media/2013Oct/0014.html
<scribe> Scribe: joesteele
Agenda

<paulc> Agenda: http://lists.w3.org/Archives/Public/public-html-media/2013Oct/0014.html
<paulc> David, are you rejoining by phone?
previous minutes

<paulc> http://www.w3.org/2013/10/01-html-media-minutes.html
paulc: fairly detailed - skipping ACTION items since there is just one
EME status and bugs

paulc: current draft has not been updated since Sept 17th?
<paulc> Current editor's draft: http://dvcs.w3.org/hg/html-media/raw-file/tip/encrypted-media/encrypted-media.html
adrianba: updated this morning
<paulc> Curent bug count: http://tinyurl.com/7tfambo
paulc: as of yesterday 20 bugs open
adrianba: resolved two bugs today
Davids Open Bug Summary

<paulc> http://lists.w3.org/Archives/Public/public-html-media/2013Oct/0001.html
paulc: going to suggest we go down through the agenda and look at what we did last time
... those actions may not have been done
Bugs on the goal section

paulc: 3 filed at the time of the CFC
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23340
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23341
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=23342
<ddorwin> having phone problems....
paulc: short discussion of this last time
<paulc> At the Oct 1 meeting the suggestion was to remove the Goals text that these bugs are referring to. Next steps?
paulc: seemed to be some discussion of taking away the offending text in the goal section
... want to return to this now?
... should we assign to someone to deal with these three bugs?
adrianba: reading the bug -- think they have the same resolution
... someone asking for use cases
paulc: so you are saying 23341 and 23342
<paulc> 23341 and 23342 are about the Goals while 23340 is about the Intro
paulc: how to deal with these?
... assign to to someone, proposed action
adrianba: proposal was made to remove the goal section -- is there consensus?
glenn: proposed we remove the goals section
<paulc> Glenn supports removing the Goals section.
<pal> I second
paulc: anyone who can't live with that?
... then proposed to resolve 23341 and 23342 by removing the goals section and ask the editors to take care of it
... do we want to do anything about the intro related bug?
... noone from task force has responded
pal: if we remove that sentence - to the end that might resolve this
<paulc> See https://dvcs.w3.org/hg/html-media/raw-file/tip/encrypted-media/encrypted-media.html#introduction
pal: could remove or change to read "this is an example flow"
<paulc> "This is just an example flow and is not intended to show all possible communication or uses."
pal: removing probably does not hurt much
... figure below just illustrates the API
<ddorwin> How about: "This is just an example flow; other combinations of APIs and events are possible."
paulc: wonder if the right way to do this is to state factually what is there?
<paulc> This diagram shows some of the interactions between the components making up a EME implementation.
glenn: figure itself probably needs some updating
<adrianba> I have removed the goals section and resolved 23341 and 23342
ddorwin: will do that once other items have stabilized
paulc: this is not meant to be proscriptive right?
<paulc> What about "This diagram shows an example flow: other combinations of API calls and events are possible."
paulc: what about this?
<ddorwin> +1
ddorwin: sounds good to me
joesteele: +1
paulc: proposal is to resolve 23340 with that text
... should remind the person in the bug that this text is non-normative
ddorwin: they are worried that this is not showing all messages
paulc: editors will respond to this then
... want to go over the tasks from last meeting
Bugs from last meeting

<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=21854
<paulc> Next step: Update spec with resolution for issues 1 and 2.
paulc: Adrian is this what you actually did?
<adrianba> I have resolved 23340
<paulc> Marked as RESOLVED FIXED. See https://www.w3.org/Bugs/Public/show_bug.cgi?id=21854#c22
adrianba: was actually three small simple changes to remove the keyadded event
... the step that added it, the mention and the table entry
subtopic: bug 19809
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=19809
<paulc> Next step: David will re-evaluate 19809 after Ade's changes 21854 are completed.
paulc: still pending?
ddorwin: yes will take care of now
paulc: need more help from TF?
ddorwin: will read through updated text and see if anything is relevant
subtopic: bug 17202
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17202
<paulc> ACTION-40?
<trackbot> ACTION-40 -- Joe Steele to Propose text for bug 17202 to propose how to share keys without leakage of information -- due 2013-10-15 -- OPEN
<trackbot> http://www.w3.org/html/wg/media/track/actions/40
joesteele: not completed
<paulc> ACTION-40 is due Oct 29
<adrianba> ACTION-40 due oct 29
<trackbot> Set ACTION-40 Propose text for bug 17202 to propose how to share keys without leakage of information due date to 2013-10-29.
paulc: last item from last time
Davids Open Bug Summary (part 2)

<paulc> http://lists.w3.org/Archives/Public/public-html-media/2013Oct/0001.html
ddorwin: 17682 we did talk about and just need to resolve
<paulc> 17682 P2 Clear Key: Document how keys and key IDs are associated
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17862 is in David's hands for implementation
https://www.w3.org/Bugs/Public/show_bug.cgi?id=17682
subtopic: bug 17750
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17750
paulc: you and Adrian were talking about this, what is this status now?
adrianba: on this, I updated the link when I was editing this morning, rest is for David
... discussion in the bug, arriving at some agreement
... some interrelated bugs here
paulc: nothing listed as dependents
adrianba: graph would be a cycle probably
... other bugs were 21855, 17199
paulc: 21855 is on David list
... can we take action on 17750 while waiting for action? or do we need to wait?
adrianba: think the status of 21855 may be slightly different than it was two weeks ago
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=21855#c9 is Adrian's latest update.
paulc: do you want to discuss now?
adrianba: yes
<ddorwin> original proposal: https://www.w3.org/Bugs/Public/show_bug.cgi?id=21855#c7
paulc: is there an explicit change you are proposing?
adrianba: think that what is in the spec explicitly supports what I described
paulc: do you want to make a more detailed proposal
adrianba: everything normative is in the spec today. discussion in the bug is about the events fired for a shared key
... how does application know which events were related to a shared key?
... two session related to a shared key - not sure which gets the update call
... thats a complication we don't need to solve for
... think we would need to solve for that then
ddorwin: think there is some difference betwene the type of shared keys being discussed
... joes key is shared key between different versions of the application
... we were thinking about sharing keys between the video and audio streams
... difference in times
... firing a key message should be up to the CDM
joesteele: I agree that there are two types of keys
<glenn> yes
joesteele: ephemeral keys like video versus audio for thes application
... and long lived keys like domain and machine keys shared between applications
... to adrians points -- I am not looking for text that explicitly supports shared keys, rather that we don't have text that explicitly prevents it
glenn: my question is related to the WebCrypto key discovery spec -- provides a mechanism for sharing machine keys, might not handle domain keys
paulc: orthogonal to this bug?
glenn: posing that question to Joe
<ddorwin> I think we should have non-normative text that the CDM should ensure all sessions appear distinct (in events/states) regardless of the underlying implementation, and maybe we should call out the CREATED-to-READY transition is possible since most of the spec uses/describes the flow that includes keymessage/PENDING.
joesteele: thinking that robustness would come into play here -- would have to review the spec to make sure
ddorwin: we may also want to call out in non-normative text that applications need to handle the ready state
<Zakim> ddorwin, you wanted to note that I think we should have non-normative text that the CDM should ensure all sessions appear distinct (in events/states) regardless of the underlying
paulc: do we need a concrete proposal for 21855?
<paulc> Did we need a concrete change proposal for https://www.w3.org/Bugs/Public/show_bug.cgi?id=21855
adrianba: think David made that proposal
joesteele: think I am ok with that text
<paulc> Editors are asked to resolve 21855 with David's proposed text.
joesteele: since it is non-normative
paulc: back to 17750 or 17919?
subtopic: bug 17750 again
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17750#c12
paulc: you and David had come to some agreement
adrianba: some dependency on 21855 because if we want to solve for explicit sharing of keys
... that would impact the text about flow
... since we are not doing that, simplifies the resolution of this bug
... prev. I argued that we should not have close unless we had a use case for it
... Joe proposed that we would like to retain the close and be a hint to the implementation
... Mark proposed a related solution as a piece to key release
... I made a new proposal to add a CLOSED state
<paulc> Mark's related proposal is in https://www.w3.org/Bugs/Public/show_bug.cgi?id=17199#c17
adrianba: and have the closed method b esomething that the application can use to ask to move to the closed state
paulc: so this is ready to be dealt with?
adrianba: think David and I are on the same page
... waiting for group consensus
<paulc> See https://www.w3.org/Bugs/Public/show_bug.cgi?id=17750#c8 for Ade's proposal.
paulc: motion to resolve using text in comment #8
<johnsim> +1
joesteele: +1
<ddorwin> +1
<paulc> In addition the c8 proposal will give us what we need to resolves 17199
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17199
paulc: we will resolve 17750 as proposed
... see if this produces a solution for 17919 as well
... on to waiting for action
<adrianba> ACTION: adrianba to implement proposal in bug 17750 comment #8 [recorded in http://www.w3.org/2013/10/15-html-media-minutes.html#action01]
<trackbot> Created ACTION-41 - Implement proposal in bug 17750 comment #8 [on Adrian Bateman - due 2013-10-22].
subtopic: bug 17673
<paulc> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17673
paulc: David had a note Sept 30 - still waiting for action?
ddorwin: are you saying we would have a CENC section and other BMFF section?
<paulc> See John's update at https://www.w3.org/Bugs/Public/show_bug.cgi?id=17673#c30
johnsim: we would have a CENC section and either add other sections as needed or add a generalized BMFF section
... at this point retaining the general formalization seems like a good way to move forward
ddorwin: this seems to leave the door open but contradicts the idea that we know about no other schemes
johnsim: the general formula that David Singer was proposing could be included
... another approach would be to remove all but CENC from the spec
... seems like folks on the call think specifying CENC should be specifying PSSH boxes
... adding too much complexity by making it general
... it is certain that we should have a CENC section as simple as I described
... question is whtether we need the more general section or just add new sections as demand is created
ddorwin: if we say that CENC passed back PSSH and others pass back SINF box then applications that get this won't know how to interpret the data
paulc: propose we queue up for next time, John has put in a proposal
johnsim: David you and I can exchange email on this
<adrianba> https://www.w3.org/Bugs/Public/show_bug.cgi?id=20991
adrianba: re: bug 20991
<adrianba> MediaKeys constructor: make CDM loading asynchronous and fix the load failure algorithm
adrianba: about loading the CDM asynchronously
... spec text as it was was broken - made an attempt at fixing
... think my change is better than what is there, but please review
... make suggestions as needed
paulc: marked as REOPENED
<ddorwin> all: Please review the event name change proposal I sent in email yesterday.
<ddorwin> http://lists.w3.org/Archives/Public/public-html-media/2013Oct/0015.html
paulc: lots of progress today
... any other business?
Summary of Action Items

[NEW] ACTION: adrianba to implement proposal in bug 17750 comment #8 [recorded in http://www.w3.org/2013/10/15-html-media-minutes.html#action01]
 
[End of minutes]
Minutes formatted by David Booth's scribe.perl version 1.138 (CVS log)
$Date: 2013-10-15 16:10:47 $