W3C home > Mailing lists > Public > public-media-fragment@w3.org > December 2009

Re: !Last phonecall before publication (be part of it)! - Media Fragments Working Group: Agenda 16 December, Telecon 1000 UTC

From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
Date: Tue, 15 Dec 2009 08:23:54 +1100
Message-ID: <2c0e02830912141323w2c6f6e72q5a6c8a10c316395a@mail.gmail.com>
To: erik mannens <erik.mannens@ugent.be>
Cc: public-media-fragment@w3.org, Guillaume Olivrin <golivrin@meraka.org.za>
Hi all,

On Mon, Dec 14, 2009 at 11:32 PM, erik mannens <erik.mannens@ugent.be> wrote:
>
> 2.1 Media Fragment URI syntax: (Yves)
> * Bug in the npt specification found by Philip:
> http://lists.w3.org/Archives/Public/public-media-fragment/2009Nov/0023.html
>
> "I'd also like to point out that the ABNF for NPT does not match
> RFC2326. It allows there to be 0 or more digits following the decimal
> point, while the MF ABNF allows 1 or more digits."
>
> ACTION to take?

I agree with Philip. I am also a bit worried about the "s" at the end
of the sec timespec for NPT, but if we disallow that on the wire, then
that's ok.

Incidentally - there are a few ABNF specs to write for the new section
5 - if somebody is keen before tomorrow...


> 2.3 Media Fragment Processing:
> * Question of MF URI validity by Philip:
> http://lists.w3.org/Archives/Public/public-media-fragment/2009Nov/0023.html
> * Suggestion of nasty test cases by Philip:
> http://lists.w3.org/Archives/Public/public-media-fragment/2009Dec/0015.html

I think I've addressed most of these, though of course the error case
section still requires a lot of work - Michael has had a good start at
that.


> 2.4 Discovery of 'Track' and 'Named' fragments:
> * ISSUE-4 [Silvia]: Should we pre-define some track names?
> * Davy's strawman implementation using ROE:
> http://lists.w3.org/Archives/Public/public-media-fragment/2009Nov/0014.html
> * Silvia's blog post:
> http://blog.gingertech.net/2009/11/25/manifests-exposing-structure-of-a-comp
> osite-media-resource/
> * Jack's proposal to write up that we should distinguish the mechanism (ROE,
> MPEG-21) vs the semantics

I still firmly believe that the actual discovery of track names and
named fragments is outside the scope of the document. All we can do is
provide examples for how it works with some existing file formats.

Happy for someone to add a paragraph on this - how to resolve the
different dimensions has not really been described yet in section 5.


> 3.5 Handling of 'Track' and 'ID':
> Conrad's proposal (Fragment header):
> http://www.w3.org/2008/WebVideo/Fragments/wiki/Server-parsed_Fragments

My section 5 has a different proposal, since everything is eventually
mapped to byte ranges. If you read section 5, you will see.


Cheers,
Silvia.
Received on Monday, 14 December 2009 21:24:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 21 September 2011 12:13:35 GMT