W3C home > Mailing lists > Public > public-i18n-cjk@w3.org > January to March 2011

RE: [css3-fonts] font selection for Unicode Variation Selector

From: Koji Ishii <kojiishi@gluesoft.co.jp>
Date: Mon, 21 Feb 2011 03:09:15 -0500
To: John Daggett <jdaggett@mozilla.com>
CC: "www-style@w3.org" <www-style@w3.org>, John Hudson <tiro@tiro.com>, "CJK discussion (public-i18n-cjk@w3.org)" <public-i18n-cjk@w3.org>
Message-ID: <A592E245B36A8949BDB0A302B375FB4E0AB201D2FC@MAILR001.mail.lan>
Thank you for the reply. I forgot to include public-i18-cjk, so cc'ed.

I have two perspectives in mind.

Theoretically, you're right, IVS's uniqueness is managed, so every font can support every variant.

In reality, it's not easy to support every variant in a single font and to catch up as IVD is updated. At a point of time, the support for IVD will vary by fonts; some fonts have this set of IVD while other fonts have other sets.

It's a little similar to what happened to Unicode decades ago; some fonts are CJK while some are Arabic. I remember there was an idea to build a global font that has every Unicode glyph, but it didn't happen. Since it was the font fallback mechanism that helped such situation in the past, people are expecting it to do so this time too.

A discussion in Japan felt me that there's a gap here, where authors are expecting IVS to change the font fallback, while doing so isn't technically easy and will not happen without explicitly writing additional spec/code, that's why I started this discussion.

The other perspective is, there's a discussion to extend the IVS/IVD idea not limited to ideographic characters but to include punctuation characters. There are some code points that were unified in Unicode and people have been in real troubles for decades. Applications developed their own method to resolve the issue; MS Word/OOXML solves this in one way, while we're discussing another way to solve the same issue in HTML/CSS or EPUB. IVS could be the right solution rather than an additional property in every application. If that goes, we might have an issue of IVS fallback between Latin fonts and CJK fonts as well, but this isn't a real problem yet.

I understand this is probably technically difficult, and I also understand that current IVS spec allows not doing this as John Hudson pointed out. But I wish you aware of that people's expectation is different.


Regards,
Koji

-----Original Message-----
From: John Daggett [mailto:jdaggett@mozilla.com] 
Sent: Monday, February 21, 2011 3:52 PM
To: Koji Ishii
Cc: www-style@w3.org; John Hudson
Subject: Re: [css3-fonts] font selection for Unicode Variation Selector

Koji Ishii wrote:

> I understand what you said from technical perspective. But from 
> authors perspective, if the author has one font for
> Adobe-Japan1 collection and another for Hanyo-Denshi collection[1], it 
> means that the author must apply appropriate font for each character 
> to render IVS correctly.

Are you thinking about a real scenario or just theoretically? 
These aren't separate collections, they reflect revisions of the IVS database to include additions. Theoretically, a font vendor could support both of these.

Regards,

John Daggett
Received on Monday, 21 February 2011 08:10:16 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 21 February 2011 08:10:16 GMT