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

Re: Yet Another Scroll-Up Idea (YASUI) in VTT

From: David Singer <singer@apple.com>
Date: Thu, 06 Sep 2012 16:11:08 -0700
Message-id: <8FF00797-17EB-4069-A993-89B9F26E1C9A@apple.com>
To: public-texttracks <public-texttracks@w3.org>

On Sep 6, 2012, at 16:02 , Simon Pieters <simonp@opera.com> wrote:

> On Fri, 07 Sep 2012 00:08:18 +0200, David Singer <singer@apple.com> wrote:
> 
>>> Woah! I was unaware of that wiki page.
>>> 
>>> Why can't such UAs just support the subset of CSS that they want to support, like 'color' and 'background-color'?
>> 
>> I think the amount of work to support general CSS parsing and so on is not insignificant.  It's like trying to pull one thread of spaghetti off your plate…you very easily end up with all of it in your lap :-(
> 
> I'm not convinced that's true. Things that want to support a limited subset of HTML can be expected to implement all of the parsing rules (or use an off-the-shelf HTML parser) and then implement only the stuff they care about without ending up with implementing a full browser rendering engine. It might be the case the off-the-shelf CSS parsers aren't readily available yet, though.
> 
> Even if I'm wrong, for WebVTT the set *allowed* CSS properties is very small.

I am aware of implementations that don't have a CSS built-in, and I am very hesitant to make a change that mandates it unless absolutely needed (and it doesn't seem to be here).

>>> The WebVTT rendering rules don't really work like "divs" at all.
>> 
>> But regions could work pretty much like them (in a restricted sense).
>> 
>>> The rendering rules do things like avoid overlapping. What happens if a rollup cue and a positioned cue within a region overlap?
>> 
>> 
>> At the moment, what I wrote was that when text is addressed into line N of a roll-up region, then the (each) line of arriving text causes the contents of lines N and above to move up a line; can collision still occur?
> 
> This assumes that you can only position cues vertically with integers, not with %?

oh, right.  % needs thought.  I will re-read the existing collision rules (maybe others can pitch in here).  But perhaps we can say that roll-up regions must be addressed by line (which is only sort-of a cop-out).

David Singer
Multimedia and Software Standards, Apple Inc.
Received on Thursday, 6 September 2012 23:11:36 UTC

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