Media Fragments Working Group Teleconference

19 Jan 2011


See also: IRC log


raphael, Yves, +329331aaaa, Erik, Davy, +31.20.592.aabb, Jack
Thomas, Silvia
Erik, Raphael


<trackbot> Date: 19 January 2011

<raphael> Regets: Thomas, Silvia

1. Admin


<raphael> scribe: yves

<raphael> scribenick: Yves

approval of last week minutes

<raphael> +1

<Yves> +1

<davy> +1

RESOLUTION: minutes approved http://www.w3.org/2011/01/12-mediafrag-minutes.html


Philip sent some patches that needs to be incorporated in the spec

As we did lots of changes, we need another LC (that will be short).

I would like that we decide on the LC next week

aiming for 3 weeks review time

any objection with the plan?

<erik> no

<hackerjack> no

Davy: do I have to include RTSP in the spec before next week?

Raphael: yes

Davy: I will try

document changed a lot in december but not a lot since then

we need to rephrase slightly the part of the spec about SMPTE timecodes

also new video codec don't have fixed framerates, we need to address that

jack: 99.99% of existing movies files will have fixed framerate even if the format allows dynamic framerate

raphael: we need to address this case

Jack: starting to rethink my position on that issue, it makes sense to present the whole fragment in that case

should it be implementation notes?

Raphael: in the final document, the things that we may remove because of lack of implementation might go in a non-normative appendix

Yves: fine as long as it is clear that it's not normative and there because of lack of implementation

Davy: variable framerate is usually not implemented as the encoder needs to communicate with the packager, and not sure it's implemented now. Also SMPTE time codes at a frame-basis should be there

Jack: how about implemetation ?

Davy: might try to get one at least partial

Jack: if people are interested in that, they should demonstrate their interest with code, even if it's a toy implementation

Raphael: would like to invite Addison to discuss issue 17 about IRI and track names

would feb 2 be ok? (ie: get critical mass here)

<raphael> I18N discussion: http://lists.w3.org/Archives/Public/public-media-fragment/2010Jun/0056.html

raphael: we also need to dedicate a complete telcon on test cases.

erik: will issue 19 be solved by next week?

raphael: yes

HTML WG liaison

<raphael> Bug: http://www.w3.org/Bugs/Public/show_bug.cgi?id=10723

Bug was rejected because it was multiple things in one bug, Silvia had an AI to open several bugs

<raphael> I have been dropped :-(

Yves: only the last one about having to stop at the end of the fragment needs reopening


action 204 will be dropped

<trackbot> Sorry, couldn't find user - 204

the header syntax should be checked in the tests

Open issues

RTSP will be added soon

test cases

Davy has some updates on that front

<raphael> drop ACTION-204

<davy> http://lists.w3.org/Archives/Public/public-media-fragment/2011Jan/0025.html

<davy> http://www.w3.org/2008/WebVideo/Fragments/TC/server-test-cases

<davy> http://www.w3.org/2008/WebVideo/Fragments/TC/server-test-cases

<davy> http://www.w3.org/2008/WebVideo/Fragments/TC/ua-test-cases

raphael: wondering about server-side tests

davy: we need extra test cases, yes
... but we have the template for test cases

Raphael: there are two things to check, URI parsed as it should and HTTP request sent as it should, and the second point is to verify the behaviour (graphical)

jack: if we have a client saving the file, we can do automatic testing

Yves: but there are things that we can't test that way (like displaying the complete timeline)

Raphael: a few things needs to check manually, which is fine

Jack: we need to check what SVG or CSS people are doing to test graphical output
... also if the number of manually tests is small...

s/manually tests/manually verifiable tests/

Jack: the first thing is to review all the current tests

Davy: we have also an issue with sources for tests

Raphael: can we make files or is transcoding too difficult?

Jack: ffmpeg might help there

Raphael: please ask on the ML.

<hackerjack> yes


<raphael> One question: should we use one of the movies in the spec and in the test cases ?

Raphael: should we use in the test cases one of the movies we have in the agenda? (ie: one with the right license)

<raphael> * Sintel: http://www.youtube.com/watch?v=ac7KhViaVqc

<raphael> ** longer version, http://www.youtube.com/watch?v=eRsGyueVLvQ

<raphael> * Big Bunny: http://www.youtube.com/watch?v=XSGBVzeBUbk

<raphael> * The Elephant Dreams: http://www.archive.org/details/ElephantsDream

<hackerjack> I would prefer not to use the suggested movies

<davy> http://www.w3.org/2008/WebVideo/Fragments/media/spatial_30fps.webm ?

Jack: we should create synthetic movies, easier to do checking (like switching backgrounds on boundaries we are about to tests)

<davy> http://ninsuna.elis.ugent.be/Media/MFWG/TC/spatial_30fps.mp4?track=3;1

erik: should we drop all the AI assigned to Michael?

raphael: safe ot drop them

Erik: we need to talk about use cases as well (from Shiraishi San)

<Nobu> Thank you, I am on IRC now.

Erik: also who is attending the workshop in Berlin?

Raphael: no

Erik: I should be there


<Nobu> I

<raphael> Nobu: we agree to continue the discussion on your use case on the mailing list

<raphael> please, create a wiki page if you want

<Nobu> I see, thanks.

<raphael> The Use Cases and Requirements document is sort of "on hold" ... which means, we will update it at the end of the rec track process

<raphael> * Sintel: http://www.youtube.com/watch?v=ac7KhViaVqc

<raphael> ** longer version, http://www.youtube.com/watch?v=eRsGyueVLvQ

<raphael> * Big Bunny: http://www.youtube.com/watch?v=XSGBVzeBUbk

<raphael> * The Elephant Dreams: http://www.archive.org/details/ElephantsDream

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2011/01/19 11:16:47 $