Re: Chromatic font research

From: Sampo Syreeni (
Date: Thu Jun 27 2002 - 06:31:24 EDT

On Wed, 26 Jun 2002 wrote:

>Sure, pictures have colour, but pictures are not characters. Not even
>pictures of things that represent characters.

Depends on what you consider a character. In my inexpert opinion, any sign
which can be used in a way reminiscent of a character (i.e. can be used to
encode an abstract concept, is used as a part of a wider, orderly system
of similar signs to encode information, and so on) *is* a character. I
mean, if it walks like a duck...

The fact that people didn't happen to have photographically accurate
megacolor printers when they first came up with the idea of writing is a
historical detail which I cannot see having anything to do with the
definition of a "character", per se.

>We could invent a system that uses 4 x 4 matrices of squares each in one
>of 16 colours with each configuration representing a different character
>from a repertoire of up to 256 characters. Should these representations
>of characters be themselves encoded as characters? No, IMO.

I'm not quite sure about this, either. Precisely such a scheme is
currently used to encode certain high capacity 2D barcodes. IMO, what
we're dealing with, here, are indeed characters. Much like the many purely
technical symbols already encoded, Braille, Dingbats, bullets and certain
other kinds of punctuation, arrows, control pictures, character
recognition and block drawing symbols, block elements, geometric shapes,
the non-Han symbols used in CJK text, musical symbols and tags. In fact,
many of those seem far less like characters than discrete barcode symbols
used to encode alphanumeric information -- many of the symbols already
encoded are not used as part of an orderly writing system, at all, but in

I think what counts is use in interchange. If there is a wide-spread need
to interchange data containing a specific symbol it should be standardized
and encoded in Unicode regardless of its genesis or similarity to older,
more conventional characters.

Sampo Syreeni, aka decoy -, tel:+358-50-5756111
student/math+cs/helsinki university,
openpgp: 050985C2/025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2

This archive was generated by hypermail 2.1.2 : Thu Jun 27 2002 - 04:51:35 EDT