[whatwg] On the subtitle format for HTML5

2010/5/25 Tab Atkins Jr. <jackalmage at gmail.com>:
> 2010/5/23 Silvia Pfeiffer <silviapfeiffer1 at gmail.com>:
>> I just came across this thread
>> http://forum.doom9.org/showthread.php?p=1397067 and found it a most
>> interesting read!
>> Particularly the comment of jiifurusu .
>>
>> It seems the subtitling community is developing a replacement format
>> for ASS with advanced features beyond what WebSRT has. Wouldn't that
>> show there is a need for an exchange format with advanced features?
>
> Not necessarily. ?It means that people want certain advanced features.
> ?It doesn't mean that those are necessary, or that the people
> developing those advanced features are aware of existing work they can
> build on, like the entire web stack. ?We can do a lot with a very
> simple format that covers all the *necessary* use-cases and can be
> easily implemented by simple devices, and then expose extra
> functionality via the web stack's technologies like CSS for the more
> important devices (that is, anything that can implement the web).
>
> This does presuppose a particular segmentation of device
> needs/priorities, but it's a segmentation that I believe makes the
> most sense for a modern format, given the reality and increasing
> pervasiveness of web-based video.
>
>
>> That new format seems to try and cater for high-end needs and lower
>> end needs. If we have to develop a new non-HTML-like format, wouldn't
>> it make sense to coordinate with those guys? In particular if the
>> community that we are trying to build upon by reusing SRT is actually
>> against extending SRT?
>
> Based on that thread, the main argument that community has against
> extending SRT is that it won't be compatible with current authoring
> tools. ?Their advice appears to be to instead adopt a new format being
> created which will also be incompatible with current authoring tools,
> though, so I don't know if I can trust their instincts too much. ?^_^
>
> (Not saying anything in particular against ASS or AS6, as I haven't
> looked at them in any sort of detail, but they do similarly appear to
> be more complex than we want for the same reasons that everything else
> has been too complex - they build in things that are potentially
> desirable but not necessary, and which can be done through existing
> web-stack technology equally well.)

The complexity argument will similarly be used by the subtitling
community that if we require all of HTML as a format for high-end
subtitling, we are bringing too much complexity into the subtitling
world.

I think we have to be careful not to make a short-sighted decision
right now with what we think is the 80% use case, which in the future
may turn into more of a 40% use case and the high end features -
things such as animations in subtitles, SVG images in subtitles,
hyperlinks in subtitles, transparent overlay images in subtitles, etc
- will be much more common, because the world's technology has moved
on and subtitles are much more common. I look in particular at what is
already possible on YouTube with subtitle-like technology such as
annotations and even the overlay ads. I know that much of this is not
for accessibility, but why would we not think beyond accessibility for
something as important as timed text for video.

Cheers,
Silvia.

Received on Monday, 24 May 2010 15:40:55 UTC