W3C home > Mailing lists > Public > public-texttracks@w3.org > September 2012

Re: Metadata in the VTT file header (bug 15851), use cases (and a need to close this)

From: Philip Jägenstedt <philipj@opera.com>
Date: Thu, 13 Sep 2012 09:00:08 +0200
To: public-texttracks <public-texttracks@w3.org>, "David Singer" <singer@apple.com>
Message-ID: <op.wkka2ilhsr6mfa@kirk>
On Wed, 12 Sep 2012 23:00:53 +0200, David Singer <singer@apple.com> wrote:

>
> On Sep 12, 2012, at 13:24 , Ian Hickson <ian@hixie.ch> wrote:
>
>> On Wed, 12 Sep 2012, David Singer wrote:
>>>
>>> Even in the case where final deployment is on the web, we identified at
>>> least one instance where keeping a file 'tagged' in workflows would be
>>> beneficial.
>>
>> (Is there a bug filed on this case, by the way?)
>
>
> I never did describe a captioning workflow, did I?  My apologies.
>
> (I was surprised when I learned this;  you may be, too.)
>
>
> Apparently, captioning and sub-titling are often done under contract by  
> other houses than the maker of the original material (not a surprise) --  
> and the original material owner (e.g. film studio) often has limited  
> rights to use the caption file(s) -- limited in both where they can use  
> the file (e.g. on TV only, not on the web, or in the USA only, not  
> Europe) and when or for how long.  The original studio often doesn't  
> (often, is not allowed to) retain the caption file after the time runs  
> out.
>
> As a result, when (enter the 3rd protagonist) wants to use a piece of  
> content and have captions, they often have to contact both the studio  
> and (if they can find out who it is) the caption house.  Confirming that  
> the caption file is correct for the version of the film they have (the  
> same edit with the same timing) is sometimes needed, and so on.
>
> (Yes, it's a mess.  No, I can't change it.)
>
>
> The upshot is that it's pretty useful for caption files to be  
> self-documenting; I don't think we need to say what attributes/values  
> they need in them (e.g. a content ID, title, etc.), but leaving space  
> for attribute value pairs that can be used for this is pretty useful.  I  
> think we proposed stating that X- attribute names were open for  
> experimentation and this kind of 'in house' use, and would be sure not  
> to conflict with a standard name defined later;  this seems enough for  
> now, to me.
>
> (Yes, you could insist that they are zipped up with their documentation,  
> but self-describing files are more clean, and -- importantly -- less  
> likely to get detached from their descriptions).
>
> Yes, this could be covered by comments, but I have seen too many places  
> where comments ended up having to have a 'required structure' to be  
> machine parsable; I'd rather reserve comments for human reading, and  
> attributes for machine reading, and not mix the two if possible.

This is a decidedly non-Web use case. Do we have any reason to believe  
that WebVTT will be used as the exchange format between film studios and  
captioning houses? I don't really see why we should try to accommodate for  
non-Web users, in particular until they themselves come to ask for it.

As for "X-":  
http://tools.ietf.org/html/draft-saintandre-xdash-considered-harmful-01

-- 
Philip Jägenstedt
Core Developer
Opera Software
Received on Thursday, 13 September 2012 07:00:48 UTC

This archive was generated by hypermail 2.3.1 : Thursday, 8 May 2014 13:18:52 UTC