Re: Bopomofo summary updated

On Thu, Feb 5, 2015 at 5:21 AM, Edward O'Connor <eoconnor@apple.com> wrote:

> Hi Richard,
>
> You wrote:
>
> > i updated the text of
> >
> > http://rishida.net/scripts/bopomofo/ontheweb
> >
> > In particular, i added a section at the bottom summarising the point to
> > which I think we may be converging.
> >
> > http://rishida.net/scripts/bopomofo/ontheweb#forward
>
> This is really useful; thanks! In your conclusion, you say that
>
> > The positioning of tones 2-4 should be achieved using OpenType font
> > rules[…]
>
> I disagree. As you go on to say,
>

> > I'm not aware of any such font at the moment
>

But creating and distributing a font which includes only bopomofo character
and can position tone mark correctly is not hard.

Browser engines need to be able to do the right thing for tone marks
> whether or not the font had such info in it. Also, tone mark positioning
> isn't like other optional font features like ligatures. Their position
> isn't stylistic.
>
> Have you tried WebKit's implementation of this recently? I'd love to get
> feedback on our tone mark positioning.
>

I just tried the WebKit Nightly today. It doesn't seem to position the tone
mark in a separate column, even with ruby-position: inter-character. I
don't see any bug in WebKit Bugzilla related to tone mark positioning of
bopomofo, either. Your bopomofo tests are still using nested ruby markup,
which should be deprecated now.

I said "even" because actually the tone mark should be positioned the same
way whether it is in ruby annotation or not. It is also one of the main
reasons I suggest using font metrics instead of implementing it in browser
engine.

- Xidorn