- From: Pierre-Anthony Lemieux <pal@sandflow.com>
- Date: Tue, 29 Apr 2014 15:46:20 -0700
- To: David Singer <singer@apple.com>
- Cc: Paul Cotton <Paul.Cotton@microsoft.com>, "public-html-media@w3.org" <public-html-media@w3.org>
Hi David, > The MP4 file format discussion list. Do you mean mp4-sys, or something else? Just trying to the reference right so that I can understand the context of and provide input on "but on the mpeg list there was exactly the opposite suggestion, that there be only one box." > Something for ‘debugging and inspection’ is quite unsuitable for being both > syntax-constrained and constrained to a fixed vocabulary. Ok. Well, I am sure a 'kind' box would have been used if one had existed back then. I am merely pointing out that there will be ISO BMFF files out there with 'kind' encoded elsewhere. Thanks, -- Pierre On Tue, Apr 29, 2014 at 9:57 AM, David Singer <singer@apple.com> wrote: > > On Apr 29, 2014, at 9:48 , Pierre-Anthony Lemieux <pal@sandflow.com> wrote: > >> Hi David, >> >>> But on the mpeg list there was exactly the opposite suggestion, >>> that there be only one box. >> >> Which MPEG list? > > The MP4 file format discussion list. > >> >>> That’s inappropriate, I think. And strange. >> >> Uh. Ok. Why? > > "name is a null-terminated string in UTF-8 characters which gives a human-readable name for the track type (for debugging and inspection purposes). “ > > It explains the type of track. Typical common practice is to put in a string from the spec.: > > ‘vide’ Video track > ‘soun’ Audio track > ‘hint’ Hint track > ‘meta’ Timed Metadata track > ‘auxv’ Auxiliary Video track > ‘text’ Timed text track > ‘subt’ Subtitle track > > Something for ‘debugging and inspection’ is quite unsuitable for being both syntax-constrained and constrained to a fixed vocabulary. > >> >> Best, >> >> -- Pierre >> >> On Tue, Apr 29, 2014 at 9:42 AM, David Singer <singer@apple.com> wrote: >>> >>> On Apr 28, 2014, at 20:27 , Pierre-Anthony Lemieux <pal@sandflow.com> wrote: >>> >>>> Hi David, >>>> >>>> Some data points/questions below. >>>> >>>>> Quantity: Zero or more >>>> >>>> Does this mean that more than one "kind" box can be associated with a >>>> single track? A given timed text track can serve multiple purposes, >>>> e.g. both SDH and subtitle, and hence be associated with multiple >>>> kinds. >>> >>> OK. But on the mpeg list there was exactly the opposite suggestion, that there be only one box. >>> >>>> >>>>> contains the track ‘kind’ as defined <<by the W3C spec // >>>>> section X of this spec. // whatever .>>. >>>> >>>> Also, the 'kind' defined in HTML might not be sufficient alone to >>>> cover common kinds of timed text out there, e.g. 'commentary', >>>> 'large'/'Text for the Visually Impaired', 'easyreader', 'karaoke', >>>> etc… >>> >>> There has to be some agreement what the kinds mean; we can’t have random communities inventing kinds and then finding they use the same word with different meanings. I think that the w3c needs to document them, and if you are not in the w3c’s spec/radar, you are on your own. >>> >>>> >>>>> Container: User data box (‘udta’) >>>> >>>> At least one specification (Ultraviolet) use the 'name' field of the >>>> 'hdlr' box to carry a comma-separated list of kinds for the timed text >>>> track. >>> >>> That’s inappropriate, I think. And strange. >>> >>>> >>>> Thanks, >>>> >>>> -- Pierre >>>> >>>> On Wed, Apr 9, 2014 at 7:15 AM, David Singer <singer@apple.com> wrote: >>>>> >>>>> On Apr 9, 2014, at 15:25 , Paul Cotton <Paul.Cotton@microsoft.com> wrote: >>>>> >>>>>> [MSE] Summary of "Resolving Bug 24370" thread and proposed next steps, Aaron >>>>>> http://lists.w3.org/Archives/Public/public-html-media/2014Apr/0043.html >>>>>> >>>>> >>>>> >>>>> I am checking with the MP4 experts at MPEG, but my inclination is that the box definition belongs in the W3C spec. anyway and would be registered at the MP4RA. Something like >>>>> >>>>> 1.1.1.1 Definition >>>>> >>>>> Box Type: ‘kind’ >>>>> Container: User data box (‘udta’) >>>>> Mandatory: No >>>>> Quantity: Zero or more >>>>> >>>>> The kind box contains the track ‘kind’ as defined <<by the W3C spec // section X of this spec. // whatever .>>. >>>>> >>>>> 1.1.1.2 Syntax >>>>> >>>>> aligned(8) class TrackKindBox >>>>> extends FullBox(‘kind’, version = 0, 0) { >>>>> string thekind; >>>>> } >>>>> >>>>> 1.1.1.3 Semantics >>>>> >>>>> thekind is a null-terminated string in UTF-8 characters, giving a track kind as defined by <<section X of Y>>. >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> David Singer >>>>> Manager, Software Standards, Apple Inc. >>>>> >>>>> >>>> >>> >>> David Singer >>> Manager, Software Standards, Apple Inc. >>> > > David Singer > Manager, Software Standards, Apple Inc. >
Received on Tuesday, 29 April 2014 22:47:08 UTC