From: philip chastney (firstname.lastname@example.org)
Date: Sun Nov 23 2008 - 06:08:32 CST
--- On Sun, 23/11/08, Hans Aberg <email@example.com> wrote:
From: Hans Aberg <firstname.lastname@example.org>
Subject: Re: Why people still want to encode precomposed letters
To: "Doug Ewell" <email@example.com>
Cc: "Unicode Mailing List" <firstname.lastname@example.org>
Date: Sunday, 23 November, 2008, 9:51 AM
On 23 Nov 2008, at 05:45, Doug Ewell wrote:
> I think Karl may have expected that fonts could be developed in such a way
that combining diacritical marks would be spaced properly above the base
character, more or less by magic. I used to think that would be possible when I
knew nothing about font design, instead of knowing almost nothing about font
design as I do now.
> I still think it would be reasonable to expect combining marks like
macrons and circumflexes to be always centered over the base character, not off
to the right, even if the vertical spacing is wrong. Like I said, almost
I experimented a bit, one and a half decade ago, with creating the Swedish
letter Å (U00C5) in TeX, which then was necessary to do by combining an A
writing a small circle above it. It turns out to be quite complicated, because
characters may have a slant, to center the circle on.
So one needs to have a more advanced font model, which for each character
contains information where to position combining characters. I am not sure about
the state of this matter - one can have Chinese fonts that combine the radicals
- so it ought to be possible to do for the combining characters as well.
as you say, combining marks are not always "centred"
how about changing the terminology, and saying combining marks are placed on the "main vertical axis"?
with roman-style letters, the main axis would normally be truly vertical, with italic-style letters the main axis might be anything from 12 to 18° off the vertical
note that, with italic-style letters, unless it's a perfect circle showing no contrast, the circle itself may also need to be slanted
This archive was generated by hypermail 2.1.5 : Sun Nov 23 2008 - 06:11:22 CST