Re: Encoding: Referring people to a list of labels

On 27/01/2014 10:20 PM, "Henri Sivonen" <hsivonen@hsivonen.fi> wrote:
>
> On Mon, Jan 27, 2014 at 11:35 AM, Andrew Cunningham
> <lang.support@gmail.com> wrote:

> If the character-to-glyph mapping of a writing system is trivial
> enough for it to be feasible to use @font-face and to commandeer e.g.
> the Latin-1 Supplement (whether encoded in UTF-8 or windows-1252) or
> to use the PUA (in UTF-8), then it is feasible to use the right code
> points with @font-face without the browser having special support for
> those code points (yes, even if the right code points are on astral
> planes).

There are a whole range of assumptions there that do not always hold true.

Legacy fonts using glyph based  encodings may involve 1 to many and many to
one relationships with Unicode codepoints. In Unicode the glyphs may need
to be reordered or repositioned or be rendered with required ligatures.

Although 8-bit encodings and pseudo-Unicode encodings don't need complex
rendering the Unicode require complex rendering. Sometimes this will
require opentype support,  sometimes it may require graphite support when
opentype support isn't sufficient at the time.

If I add the need for syllable or word boundary itentification for line
wrapping...

Most browsers still have a long way to go.

A.

> --
> Henri Sivonen
> hsivonen@hsivonen.fi
> https://hsivonen.fi/