Re: [css-text-decor] Emphasis marks and auto-hiding ruby annotation

On 11/13/2015 09:02 AM, Koji Ishii wrote:
> On 11/11/2015 11:02 AM, Xidorn Quan wrote:> Hi,
>>
>> I have a question that, if the annotation of a character is
>> autohidden, where should the emphasis mark be drawn for this
>> character?
>>
>> The current spec says:
>>  # If emphasis marks are applied to characters for which ruby is drawn
>>  # in the same position as the emphasis mark, the emphasis marks are
>>  # placed outside the ruby.
>>
>> It reads to me that if the annotation is autohidden, the emphasis mark
>> would be rendered immediately on top of the character, and in the same
>> line as other annotations, is that correct?
>>
>> To be more clear, the question is, if we have ruby like:
>>   ふ がな
>>   振り仮名
>> and we apply emphasis marks on it, which result makes more sense?
>>   、、、、
>>   ふ がな
>>   振り仮名
>> or
>>   、 、、
>>   ふ、がな
>>   振り仮名
>> ？
>
> The situation can happen without auto-hide, and I saw both examples, so
> the spec avoids saying which specifically, only "emphasis marks come
>outside of ruby" (figure 9 of the spec[1].)
>
> Authors may want to specify in future, or can live with either, I can't
>answer at this moment.
>
> In my personal opinion, especially when the middle part is missing, the
> former (consistent position) looks better. On the other hand, if emphasis
> marks are on the whole paragraph, and only one character has ruby, the
> latter might look better. Choosing either by auto-hide or not might be
>one possible option.
>
> [1] https://drafts.csswg.org/css-text-decor-3/#text-emphasis-position-property

There was previous discussion on the positioning of emphasis marks
in the presence of ruby. See
   http://www.w3.org/mid/AEB4771C-8ED0-49DD-B444-34F2BE54F85F@w3.org
     continuing at https://www.w3.org/mid/4D277EC5.2050607@w3.org
   https://www.w3.org/mid/A592E245B36A8949BDB0A302B375FB4E201BF47330@MAILR001.mail.lan
   https://www.w3.org/mid/510C6999.9050000@inkedblade.net

The conclusion was that the behavior illustrated in the spec is required,
which is why we did not illustrate both options. :) See
   https://www.w3.org/mid/5136E94C.9010605@inkedblade.net

That said, I think auto-hidden ruby is a special case, and we should
probably spec Gecko's behavior for that.

A related question is then, what about empty annotations? Should they
match the behavior of auto-hidden ruby or non-annotated text?

CCing Murakami-san for thoughts. :)

~fantasai

Received on Thursday, 22 December 2016 21:13:24 UTC