Re: [css-ruby] Should `ruby-merge' initial value be `auto'?

On 01/24/2014 09:47 AM, kawabata taichi wrote:
> Dear people concerned with CSS-Ruby,
>
> Currently, CSS Ruby `ruby-merge` property's initial value is
> `separate`.(http://www.w3.org/TR/css3-ruby/#collapsed-ruby). I would
> prefer that this initial value be `auto` instead of 'separate'.
>
> Rationale: Multiple <rb> tag within single <ruby> tag is typically
> used for compound words (Jukugo). For Jukugo, the best rendering
> algorithm is "Jukugo-Ruby algorithm", as its name says (see JLREQ
> 3.3.7). By specifying `auto` by default, the rendering engine would
> display the ruby text by the best algorithm it supports (including
> Jukugo-Ruby algorithm).
>
> The current initial value, 'separate', is typically suitable if multiple
> <rb>s do not constitute a compound word. In such a case, an author would
> put them in different ruby segments, thus this spec is a bit
> unnatural.
>
> As of it, I would like to propose to change the initial value of
> `ruby-merge' to be 'auto' instead of 'separate'.

Hi Kawabata-san,

We chose the initial value of ruby-merge as 'separate' because
this is the simplest to implement, and would result in consistent
results for implementations that support ruby-merge's advanced
behaviors and those that don't.

If we make it 'auto', then older/simpler implementations will
display as 'separate', and more sophisticated ones will display as
jukugo. This may or may not be what the author intended. So we
chose 'separate' as the initial value. This will give consistent
results, and won't surprisingly display as jukugo in some browsers
when the author has only seen 'separate' in his browser. To get
jukugo, the author has to specify 'auto' explicitly. This makes us
more sure that it is what he really wants.

At least, this was the logic I had. I'm open to counter-arguments. :)

~fantasai

Received on Monday, 27 January 2014 07:51:33 UTC