Re: [css3-text] fullwidth/halfwidth (was text-transform values should be grouped

Hi, Ishii-san, thank you for your reply.

On 2010/10/08 17:37, Koji Ishii wrote:
>>> If you have use case for "halfwidth", or other use cases for "fullwidth",
>> would you please share with us?
>>
>> How about for user style sheet? Most Japanese news sites which are owned
>> by newspaper companies are using fullwidth alphabets and numbers by some
>> reasons. But they are not in Western fonts, but some users may want to display
>> them by Western fonts which are more beautiful than Japanese fonts.
>
> That makes sense, but it introduces a new issue. If we have this use case in mind, you probably do not want "halfwidth", because it converts fullwidth Katakana to halfwidth Katakana in Unicode compatibility zone.

Oh, you're right.

> We could possibly have "normalize" or "canonicalize" instead to avoid this issue. But then, the feature becomes much more complicated to implement, and is not a styling issue any more for most of the scripts other than Japanese. It's hard to convince people saying "you can run Unicode canonicalization process when you get data from database or other sites, and you should."
>
> The other possible solution is to name it "halfwidth-except-compatibility-zone", but I think it's a little too complicated and too specific to implement in CSS.

I agree...

http://www.unicode.org/charts/PDF/UFF00.pdf

They are grouped by digits, alphabets, kana, hangle and others. If we 
add a feature like "halfwidth-except-compatibility-zone", some users 
might want other sets of the groups.

I think that taking only fullwidth may make sense, unfortunately.

-- 
Masayuki Nakano <masayuki@d-toybox.com>
Manager, Internationalization, Mozilla Japan.

Received on Monday, 11 October 2010 00:50:12 UTC