See also: IRC log
<trackbot> Date: 01 April 2009
<scribe> Scribe: Raphael
<scribe> scribenick: raphael
Propose to accept the minutes of 25/03?
RESOLUTION: minutes of 25/03 are accepted
Raphael: should we move the telecon hour, to be more convenient for Silvia ?
Jack: I will miss one telecon every two weeks
Raphael: proposal to have the
telecon 2 hours later
... to be more convenient for Silvia
<scribe> ACTION: Raphael asks Silvia if she would like to have the telecon 2 hours later [recorded in http://www.w3.org/2009/04/01-mediafrag-minutes.html#action01]
<trackbot> Sorry, couldn't find user - Raphael
<scribe> ACTION: Raphaël asks Silvia if she would like to have the telecon 2 hours later [recorded in http://www.w3.org/2009/04/01-mediafrag-minutes.html#action02]
<trackbot> Created ACTION-54 - Asks Silvia if she would like to have the telecon 2 hours later [on Raphaël Troncy - due 2009-04-08].
Erik list the participants: Michael, Jack, Erik, Davy, Raphael, Conrad, David, Franck, Guillaume
scribe: + Yves / Silvia on the phone
Erik: we will re-organize the section 5 significantly
Raphael: my proposal, finish the
document by next Wed 8 of April, and vote for publishing the
... in our next telecon
... make a quick publication request
and hope to have the document published on the 9th
Jack: Michael took my action,
because I was busy ... but perhaps I can take over
... I will send an email to Michael, to ask him if I can take over
Erik: difference between reference and informative sections?
Raphael: informative parts are just for giving hints to the reader, generally dtd or schemas
Problem 1: not use '%' but spell it
<scribe> ACTION: Yves to change the formal grammar to have 'percent' spelled [recorded in http://www.w3.org/2009/04/01-mediafrag-minutes.html#action03]
<trackbot> Created ACTION-55 - Change the formal grammar to have 'percent' spelled [on Yves Lafon - due 2009-04-08].
Problem 2: use of single quote (') and sub-delims in utf8string
<scribe> ACTION: Yves to change the formal grammar to have unreserved characters + %-escaped ones in utf8string and _not_ a set of pchars [recorded in http://www.w3.org/2009/04/01-mediafrag-minutes.html#action04]
<trackbot> Created ACTION-56 - Change the formal grammar to have unreserved characters + %-escaped ones in utf8string and _not_ a set of pchars [on Yves Lafon - due 2009-04-08].
3.2: Technologies Survey
Erik and Davy will work on it on Monday / Tuesday
3.3: HTTP implementation
<trackbot> ACTION-52 Replace all examples in the HTTP implementation wiki page with the video Jack took closed
Raphael: how can I have the byte ranges that correspond to my temporal interval?
Jack: now we are entering into a
different problem, that is, should we do transcoding or
... the bytes range will be different depending on transcoding or not
... I have just sent an email stating we should consider more the user
... see: http://lists.w3.org/Archives/Public/public-media-fragment/2009Apr/0003.html
Raphael: I think we should go for
your option 2, and define what is reasonable
... what is the point 4? Is it related to the in-context discussion?
Jack: no, no, no
Point 5: we leave that away, we haven't specified what the spatial cropping will actually do!
Point 3: it seems to be related to the issue: transcoding or not
Davy: if we want to re-syn the audio and video, then we need to transcode
Raphael: what is the worst thing that can happen?
Davy: the audio and video might not be aligned
Jack: the reason while I use a wget example, is that you want to store this piece and re-use it later on
Davy: my proposal was to do no
transcoding and have a reasonable out-of-synch
... same than for the I-Frame
... we choose the closest timeframe
Jack: what is reasonable?
Davy: I think there is studies
which show what is the largest not-synch temporal interval that
users can support
... you and I have different interpretation of what we allow for transcoding
... I prefer Jack's interpretation
Jack: I had interpreted it as "we will not standardise any type of fragmenting that we think cannot be implemented without using transcoding".
Davy: I interpret it as "an implementation is not allowed to do any transcoding".
Davy also now prefers Jack's interpretation
Jack: I think we can standardize a spec that can be implemented without transcoding
Raphael: I'm also now confused, it is different from your previous statement
Jack: if we have no boundary cases for media fragments, it should work
Raphael: should we make the algorithm in the spec, that says, jump to the previous I-Frame, etc.
Jack: no, we should have in the spec, the user expectation
Silvia: we should send larger sequence, in order to not transcode
<davy> +1 to Silvia
+1 to Silvia
Silvia: get wider intervals, on a best-effort basis
Silvia: talk about the new ogg spec
<davy> I think mp4 is also capable of skipping frames
Silvia: the YouTube player is also working on a best effort, if you click on a time offset, then the player goes to the closest frame he can start to play
Raphael: what is the closest? can it be after?
Silvia: I think it always goes
back in time
... I need to check, but this is my experience
<trackbot> ACTION-53 Check whether ffmpeg can slide movies easily, need for implementation on client and server side closed
<jackjansen> give me an action
<scribe> ACTION: Jack to enter formally in the tracker the ISSUE he has found, regarding a) transcoding and b) spatial cropping [recorded in http://www.w3.org/2009/04/01-mediafrag-minutes.html#action05]
<trackbot> Created ACTION-57 - Enter formally in the tracker the ISSUE he has found, regarding a) transcoding and b) spatial cropping [on Jack Jansen - due 2009-04-08].
Silvia: I would like to have the
telecon 2 hours later
... I think the section use case and requirements is ready
<Gui> I agree
<Gui> I am all sorry for myself, I didn't know I was late till now ! :-( I thought we were back to normal hours.
Davy: I will use our platform to
discover the mapping between seconds and bytes for this
... and update the wiki
No, AOB, Meeting adjourned