See also: IRC log
<trackbot> Date: 14 August 2012
<scribe> Scribe: Matt
<Johnsim> zakim aagg is me
<Johnsim> zakim aaaa is me
matt: Hi, I'm Matt. I'll be helping out as part of the work on Web and TV. I'm at MIT, so I'm on East Coast time.
paulc: No comments on those.
-> http://lists.w3.org/Archives/Public/public-html-media/2012Jul/0096.html Previous Minutes
paulc: There weren't any in tracker, but there were some informal ones that I put in the agenda.
-> http://dvcs.w3.org/hg/html-media/raw-file/tip/media-source/media-source.html Draft
->
&bug_id=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&cmdtype=doit&order=Reuse+same+sort+as+last+time&field0-0-0=noop&type0-0-0=noop&value0-0-0= Bugzilla
<paulc> http://lists.w3.org/Archives/Public/public-html-media/2012Aug/0003.html
-> http://lists.w3.org/Archives/Public/public-html-media/2012Jul/0003.html
/
Aaron?: I don't think there's anything to talk about they're done.
paulc: Bug 17739 and Bug 18389 are resolved
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18389 Bug 18389
<acolwell> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18550
paulc: The new bugs in item six
are in the encrypted media group and I'll deal with those at
the next meeting.
... Both of them?
<paulc> Note that bugs 18515 and 18531 are both EME bugs
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18550 New bug
acolwell: Provides a definition for TrackID
paulc: Does that propose a solution?
Aaron: Just need to figure out the text to put in there.
paulc: Timetable?
Aaron: Next call.
INFORMAL ACTION: acolwell to propose a solution for 18550 before next meeting
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=18400 Bug 18400
paulc: Previous notes said markw will address by this call.
markw: I posted a proposal to the
bug, they were descriptive rather than text proposals. No
comments yet, so next step is text proposal.
... One question: sometimes you don't know where the media ends
in a segment that has been appended. This can happen in a video
with subtitles, there's a subtitle between 5s and 10s, but then
you don't know which portion is addressed by that
segment.
... If you have the exact information about how the timeline is
covered, then you don't need that feature. There's a case of
the mp4 format ??? despite having the span, that information is
made available to the application.
<markw> regarding the mp4 format, the information about the end of the segment not in the segment, but is in the segment index
<paulc> Aaron responded in https://www.w3.org/Bugs/Public/show_bug.cgi?id=18400#c4
acolwell: For the most part I
agree with Mark on his proposal. There's only one section that
doesn't get to how long the segment is and causes a problem.
Should we mandate that media segments have some form of
duration information in them?
... I think that might be what markw was advocating, or at
least in ISO you already know it.
... If we require the duration, then we need to change the WebM
text as it's not required there.
markw: The information isn't there itself but it's available to the application. You could imagine adding to the API that that information is added when you append.
paulc: Next step?
INFORMAL ACTION: markw to reply to Aaron's response to bug 18400 comment 4 by next meeting
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17001 Bug 17001
paulc: Any action on this?
<paulc> Bob's response is here: https://www.w3.org/Bugs/Public/show_bug.cgi?id=17001#c1
acolwell: I did contact Bob, didn't CC list. He had commented back that he was okay with the changes. So I closed the bug.
RESOLUTION: bug 17001 ( https://www.w3.org/Bugs/Public/show_bug.cgi?id=17001) FIXED
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17000 Bug 17000
adrianba: We haven't worked on this, the encrypted work is a higher priority.
INFORMAL ACTION: adrianba and Johnsim to look at bug 17000 before next meeting.
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17002 Bug 17002
paulc: On this one we were trying to decide to close this bug and make a new one, adrianba said we should retitle it. acolwell said he'd look at the history.
acolwell: I retitled the bug and added a comment.
<acolwell> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17002#c7
<paulc> New title: Specify a mechanism to determine which SourceBuffer an AudioTrack,VideoTrack, or TextTrack belong to.
acolwell: Also posted a message to the list to ask for people to look at it.
-> http://lists.w3.org/Archives/Public/public-html-media/2012Aug/0008.html acolwell mail on updated bug 17002
paulc: Anyone have comments on
this now?
... It would help if people would indicate support. Please
don't think you only have to respond in the negative.
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17006 Bug 17006
acolwell: This goes back to the
DASH spec, and this asks what level of support do we want. DASH
allows language and kind to be set on each different track.
Depending on how you allow this, you could have track language
change mid playback.
... It needs to be scoped a bit better for where it is allowed
to be changed.
markw: What scenarios do you think those properties would change in DASH?
acolwell: language and role (which is kind in HTML) that could be applied to both adaptation-set? and component? Seems like you could have multiple ?? in a adaptation-set.
markw: Those are required to have
the same properties at the ??adaptation?? level.
... If you had a adaptation set with three in it, all three
would have the same values.
acolwell: I need to start a thread to discuss this. When I looked at the mpeg spec, it looked like there could be trickiness involved.
INFORMAL ACTION: acolwell to start a thread on bug 17006
<markw> the trickiness is that a single source buffer can map to multiple HTML track if the media is multiplexed
BobLund: The media component would provide an override and if you had a component that each would -- if you had an mpg2 transport with mp3, etc, and each would map into audio, video and text track.
acolwell: Do you need to be able to update each track's kind and language or groups of these? Is it at source buffer level or track level?
BobLund: I don't know if there's anything in the DASH spec that would prohibit that. We've talked about ?? and that would be a new source buffer.
markw: I'm not sure how you could match up the media components and the multiple tracks provided from a single source buffer.
BobLund: That's another issue actually.
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17072 Bug 17072
paulc: We said at the last meeting that this might not be relevant.
acolwell: Yes, most of the remaining issues are covered by what Mark is working on. Closed it, left a comment and pointed forward to new bug.
paulc: So that one is done.
acolwell: Bug 17072 was marked resolved won't fix.
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=17094 Bug 17094
Define segment formats for MPEG2-TS
paulc: acolwell You suggested this might be controversial and need to talk it out.
acolwell: No progress made. Still pending.
??: I don't know if someone on the call is championing this one, they might be best to advocate for it.
paulc: This was filed by Duncan
from BBC. Not on the call today.
... I don't see him participating in the dialog at all.
BobLund: I'll volunteer to champion this.
paulc: Perhaps do a PRO/CON list on the list for taking action on this
BobLund: Will do and will close the loop with Duncan on this.
-> https://www.w3.org/Bugs/Public/show_bug.cgi?id=16998 Bug 16998 Change sourceAppend() to take a URL and optional range parameters
paulc: We talked about splitting
this bug at the last meeting.
... It wasn't obvious if we'd made progress on that or not.
acolwell: No, I haven't split the bug.
paulc: So this one is still pending.
paulc: I have a feeling we've
touched on all of the bugs in this discussion.
... So, we've made some real progress here.
paulc: Any other business?
... Matt, will you be scribing future meetings?
matt: Attending, scribing no.
markw: What's the timeline?
paulc: If we can get the bugs
closed, there are only 7 still outstanding, publish and bring
it to the group. If we can make substantive progress over 2
weeks, and assume we don't get them all closed, then a month
from now get the remaining ones.
... That would put us mid-september, and we could put it to the
group then. They should be doing a heartbeat publication then,
and perhaps get us on the same schedule.
... I suspect we'll have to give the WG a few weeks to review
the document.
... But, our first objective is to close this round of
bugs.
... The feeling before was that we could move this and EME
forward separately.
... At the other meeting, I predicted a single digit number of
weeks away from going back to the WG. So, I think we're on
schedule, but the bar to get over is to close these last 7
bugs.
... Scribe in two weeks?
<no one>
paulc: Please clean the minutes
and send to the media list.
... Adjourned.
s|s///||
s|s/-> http://lists.w3.org/Archives/Public/public-html-media/2012Jul/0003.html/||
s|https://www.w3.org/Bugs/Public/buglist.cgi?query_format=advanced&short_desc_t.*||