W3C home > Mailing lists > Public > www-style@w3.org > June 2010

Re: [css-fonts] Is it possible to select a vertical variant in a font?

From: John Daggett <jdaggett@mozilla.com>
Date: Thu, 17 Jun 2010 00:44:41 -0700 (PDT)
To: Ishii Koji <kojiishi@gluesoft.co.jp>
Cc: John Hudson <tiro@tiro.com>, www-style@w3.org, David Lemon <lemon@adobe.com>, Ken Lunde <lunde@adobe.com>
Message-ID: <1949127012.475675.1276760681320.JavaMail.root@cm-mail03.mozilla.org>

> Will you also please tell me how I can propose to add, say,
> "font-variant-vertical" property?
> As you pointed out, vertical variant could be a little different from
> other OT features, as it is required to turn on automatically in
> vertical text flow. But in some rare cases, I would like to turn it on
> regardless of the current flow. Since it's one of the font features as
> you explained to me, "font-variant-*" looks the best place to add to
> me.
> Do you think this is a reasonable request? Since I'm still very new to
> this ML, I would appreciate to know how I can make the discussion
> progress and propose the addition of this property.

As fantasai has already written, the use of vertical variants should
probably be based on the 'writing-mode' property.  Like other
language/script sensitive features in OpenType, using vertical variants
is a *requirement* for rendering vertical text in Japanese, it's not a
stylistic choice as is the case with other proposed font-variant-xxx

How to support tatechuyoko, the display of Latin text and numerals
horizontally in vertical text runs, is a more subtle question.  This
usually involves substitution of half-width or third-width forms.  

In the example below note the rendering of '90' and '65' using
half-width glyphs and '100' using third-width glyphs:


In the same example, 1 euro = 111 yen is printed with '111' displayed
inline (i.e. not using tatechuyoko).

The XSL 1.1 spec defines one possible solution, adding the
'tb-lr-in-lr-pairs' value of the 'writing-mode' property:


I don't think this is quite the right solution, it misses the fact
that tatechuyoko is primarily used for Latin characters and
numerals (there are no half-width Kanji glyphs usually) and there's a
contextual choice of half-width/third-width forms to be made somehow,
either explicitly or inferred based on some rule.

I would emphasize again here that I think vertical text support in CSS really
needs to be considered comprehensively, I don't think it's going to work
to simply go through CSS and add a property here and tweak a property there.


John Daggett
Received on Thursday, 17 June 2010 07:45:15 UTC

This archive was generated by hypermail 2.4.0 : Monday, 23 January 2023 02:13:47 UTC