Luciano Perondi wrote:

> Ma in pratica la questione è programmare un software (un
> font) che prende il
> codice unicode ("consonant sign K", "consonant sign G",
> "consonant sign J",
> ..., "vowel sign A", "vowel sign I") e gli assegna un glifo colorato?

["But, in practice, the matter is programming a software (a font) which
takes the Unicode code (...) and assigns to it a colored glyph."]

The current OpenType specification already has a minimal support for color:
glyphs can be marked with a property indicating that the rendering software
should display them in an alternate color.

This feature (which is very rarely implemented, BTW) was added to allow
automatic coloring of Arabic vowel marks: in some calligraphic and
typesetting traditions it is customary to have black letters with red
vowels.

This feature is clearly quite ad-hoc and limited:

1) Only a whole glyph can be colored, not part of it; e.g., you
cannot have a black "j" with a red dot, because the dot is part of the "j"
glyph.

2) Only two colors are supported: the "letter color" and the "marks
color".

3) The color values (e.g., black and red) is specified *out* of the
font (e.g., in the formatting options of the word processor using the font).

Two years ago, John Hudson, with some help by me, drafted a proposal for an
extension to the OpenType specs, in order to override limitation (1).

The rationale for having a glyph with two colors was to allow the software
implementation of a feature of the Ethiopic script: in some calligraphic
traditions, some punctuation marks (namely the end-of-paragraph mark, which
normally looks like a double colon "::"), have decorative dots in a
different color (normally red or gold, if I recall correctly).

I have uploaded a sample of a decorated end-of-paragraph (taken from that
OpenType proposal) in
<http://groups.yahoo.com/group/qalam/files/colors.gif>.

I don't know what happened to John's proposal, but I assume that it was
refused, or simply forgotten. This is not surprising, as these colored dots
are purely decorative, optional, and are only used in some Coptic religious
texts, so I guess that implementing them is not exactly priority one for the
IT industry.

Anyway, implementing a font for a script where color is a structural
feature, would also require overriding limitations (2) and (3) above. But I
suspect that this would be a useless exercise, as I don't think that many
people need typing in Aztec on a computer...

_ Marco