Re: ISSUE-358 (Sequence support): Add identifier support for related ordered sequences of documents [TTML2]

On Thu, Dec 11, 2014 at 1:17 AM, Nigel Megitt <nigel.megitt@bbc.co.uk>
wrote:

>  From: Glenn Adams <glenn@skynav.com> Date: Wednesday, 10 December 2014
> 18:33
>
>   Now that I've introduced a generic metadata item element, it is better
> to satisfy this requirement by simply defining new named metadata items,
> representing it as:
>
>  <ttm:item name="sequenceIdentifier">*string*</ttm:item>
>
>  <ttm:item name="sequenceNumber">*positiveInteger*</ttm:item>
>
>
>  ttm:item isn't permitted on tt:tt though, which means that there's a
> deep parsing requirement to discover this data, so it doesn't satisfy the
> requirement.
>

yes, that's correct, it would require looking into <head/>, but that is
analogous to

<head>
<meta charset="utf-8"/>
</head>

in HTML5

also, as you might guess, I somewhat modeled <ttm:item
name="...">value</ttm:item> on HTML's well used <meta name="..."
content="value"/> construct


>
>  Incidentally, if you were to add this concept to isd:isd then the
> isd:sequence element would be unnecessary except as a root element in an
> XML document, since you could represent a sequence of ISDs using the same
> syntax.
>

an XML document must have a single root element, it cannot have a sequence
of elements at root level, so the primary purpose of isd:sequence is to
serve that function


> I see that isd:sequence and isd:isd do permit content from Metadata.class
> whereas tt:tt does not – I'm not sure why there's a difference there, in
> principle anyway.
>

since isd:sequence (and isd:isd) can potentially have document level
metadata, i.e., derived from <head><metadata>...</metadata></head>, and
since I didn't specify a head element in ISD syntax, I had to put them
somewhere, so I put them at first position among children


>
>
>
> On Wed, Dec 10, 2014 at 9:56 AM, Timed Text Working Group Issue Tracker <
> sysbot+tracker@w3.org> wrote:
>
>> ISSUE-358 (Sequence support): Add identifier support for related ordered
>> sequences of documents [TTML2]
>>
>> http://www.w3.org/AudioVideo/TT/tracker/issues/358
>>
>> Raised by: Nigel Megitt
>> On product: TTML2
>>
>> In an authoring environment that creates an ordered sequence of TTML
>> documents that collectively form some kind of coherent representation
>> collectively, such as a sequence of documents created 'live', it is useful
>> to be able to identify at the root level, i.e. with minimal parsing
>> requirements, that:
>>
>> 1. A collection of documents belongs to the same sequence and
>> 2. The relative ordering of a document within the sequence.
>>
>> This also relates to Change Proposal 25 and Issue-288, and the proposal
>> made in the November TPAC meeting.
>>
>> Modified proposal now includes a sequence number:
>>
>> Metadata attribute ttm:sequenceIdentifier
>> Considered significant on tt:tt only.
>> Optional attribute.
>> Type xs:string
>> Description: The ttm:sequenceIdentifier metadata attribute can be used to
>> distinguish between sequences of documents where the documents within a
>> sequence are related to each other but there is not necessarily any
>> relationship between different sequences.
>>
>> Metadata attribute ttm:sequenceNumber
>> Considered significant on tt:tt only.
>> Optional attribute.
>> Type xs:positiveInteger
>> Description: The ttm:sequenceNumber metadata attribute can be used to
>> sort documents with the same sequenceIdentifier into a defined order. Two
>> documents with the same sequenceIdentifier SHOULD NOT have the same
>> sequenceNumber, if sequenceNumber is used.
>>
>>
>>
>> Note that this proposal mirrors work happening in EBU but is my proposal
>> and not an EBU proposal, for the avoidance of doubt.
>>
>>
>>
>>
>

Received on Thursday, 11 December 2014 17:09:33 UTC