Re: [css3-lists] CJK numbering algorithms

On Sat, Apr 23, 2011 at 12:46 AM, Koji Ishii <kojiishi@gluesoft.co.jp>wrote:

> But what I'd like to put into CSS3 Lists spec is *not* the correct answer
> for an academic. I'd like to put the most useful and the best adaption of
> Kanji numbering style in Japanese into the CSS3 Lists spec. These two --
> academically correctness and what to put into a spec -- are the same in most
> cases, but sometimes they should be different, and I think this is one of
> the cases where we should have different answer for the spec.
>

But Korean counter style doesn't specify representations for negative
numbers as well (at least in the current draft).  And in my opinion,
negative numbering in Kanji characters itself isn't useful.

>From that view point, can we agree on option 5
> > 5. ..., マイナス参, マイナス弐, マイナス壱, 〇, 壱, 弐, 参, ...
> would be the best option to take?
>

The current spec likewise doesn't support any number above 10^16 - 1.
 Should the spec be "useful" and specify the representations for numbers
above 10^16 - 1?  In my opinion, we shouldn't because there are very rare
cases where such a large number is useful and people aren't familiar with
group markers such as 京 and 垓.

The same principle applies to negative numbers.  People rarely write
negative numbers in Kanji, and when they do, people tend to use
context-sensitive prefixes or suffixes such as 氷点下 and ▲.  And from that
point of view, any one prefix we pick will only be correct in some but not
all context and I don't think we should be spec'ing that.

- Ryosuke