RE: Chromatic font research

From: Marco Cimarosti (
Date: Wed Jun 26 2002 - 04:31:46 EDT

Peter Constable wrote:
> On 06/25/2002 11:17:52 AM Marco Cimarosti wrote:
> >Use of the Private Use Area is never questionable, as far as
> it remains
> >*private*.
> I think this might be overreacting -- or at least it can seem
> to be so. [...]

Sorry if I made this impression: I actually had a quite different thing in

I meant that, as a *minimum*, private usage of PUA is perfectly acceptable.
(I was not talking about publicly documented usages, which I don't consider
"illegal" a priori).

William O.'s idea of Christmas leaf PUA character to demonstrate a certain
technology (chromatic fonts) is *definitely* private and, hence, definitely

What I argued is that such an ad-hoc character was an inappropriate example,
because it would distract people from the main fact: that it has two colors.

I mean: people would ask: "What's that character?"; "If you wanted a
multi-color Christmas decoration, why didn't you insert a picture?"; and so
on. On the other hand, showing an entire Arabic (or English) text where all
dots have a different color from the rest of text makes it clear that you
could not use pictures to do it.

> BTW, while I think your character is perfectly fine for you
> to experiment with (and has a fun aspect to it), I am quite
> concerned at your choice of name: it suggests that it might
> be reasonable for someone to encode characters with
> specific chromatic attributes. I definitely think this
> would be inappropriate for a character set encoding standard.

There is a remote theoretical possibility that this statement may become
incorrect in the future.

In all the examples we made so far, including William's leaf, color is just
decorative so it doesn't belong to encoding itself.

However, the Aztec script uses color has a structural element: signs with
the same design can mean different things if painted in different colors.
So, if scholars *would* agree that Aztec is "writing", and if this script
*would* get into Unicode, then color *should* have to be considered also at
the encoding level.


This archive was generated by hypermail 2.1.2 : Wed Jun 26 2002 - 03:36:15 EDT