W3C home > Mailing lists > Public > public-i18n-cjk@w3.org > July to September 2010

Re: Thoughts on ruby

From: fantasai <fantasai.lists@inkedblade.net>
Date: Mon, 27 Sep 2010 19:00:44 -0700
Message-ID: <4CA14C4C.3030004@inkedblade.net>
To: Richard Ishida <ishida@w3.org>
CC: "'MURATA Makoto (FAMILY Given)'" <eb2m-mrt@asahi-net.or.jp>, public-i18n-cjk@w3.org
On 09/27/2010 07:16 AM, Richard Ishida wrote:
> 
>> Re: [3]  Some people think that nested ruby elements provide
>> ruby text on both sides.  This is syntactically allowed, but HTML5
>> says nothing about ruby text on both sides.  I do not think that
>> it is more complex than "complex ruby" of XHTML Ruby annotation.
>>
>> <ruby>
>>    <ruby>東<rt>とう</rt>南<rt>なん</rt></ruby>
>>    <rt>たつみ</rt>
>> </ruby>
> 
> Yes, this is not really so bad.  It's not so powerful as complex ruby, but
> perhaps it is good enough for supporting ruby specifically.

The structure strikes me as wrong. Structurally, this means that
たつみ is an annotation of the *annotated combination of*
<ruby>東<rt>とう</rt>南<rt>なん</rt></ruby>. That is not the intent.
The base of the たつみ is 東南 characters only.

> Elika, I'm not yet familiar with inline tables in CSS, but I've been
> wondering whether they are a better candidate for handling things like
> linguistic glosses and such.

You could style <ruby> as tables, yes, that should be pretty straightforward.
But it would affect word boundary calculations. Linguistic glosses that
span e.g. per English syllable would cause the word to break.

Ruby doesn't seem the best way to handle simultaneous translation, but
for linguistic glosses, where one text is the base and the other is
the annotation, I think it's the best option.

> I'm beginning to wonder whether we should refocus the scope of ruby markup
> and styling to specifically address just ruby needs for CJK.  The current
> model was always  vaguely intended to be useable for other things, but
> although I've heard numerous people suggest how it could be used for things
> other than ruby, I don't think I've really come across a convincing case,
> especially given the two rubys per base character limitation.

Here are some examples for you to chew on:
  http://fantasai.inkedblade.net/style/discuss/directions/scans/genji1
    The numbers atop the English phrases are effectively full-size ruby
    annotations. Note line breaking behavior and word boundaries: you
    can't get that with tables.
  http://fantasai.inkedblade.net/style/discuss/directions/scans/genji2
    Here, near the centerfold, Kanji ruby atop English base text.

~fantasai
Received on Tuesday, 28 September 2010 02:01:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 28 September 2010 02:01:26 GMT