Re: UNICODE

From: Philippe Verdy <verdy_p_at_wanadoo.fr>
Date: Wed, 4 Jul 2012 02:24:40 +0200

2012/7/3 William_J_G Overington <wjgo_10009_at_btinternet.com>:
> It is important to know that encoding is of monochrome characters.

Not really. Characters that have distinct color semantics may be
encoded distinctly with their color even if they will also need to
have a monochrome representative glyph that will remain informative
and that will not necessarily fully show in that glyph every aspect of
the character.

THe representative glyph in the charts only try to make efforts to
exhivit the distinctions, possibly using other eans such as by using
fill patterns. "White" and "Black" in charater names however do not
necessarily mean the color they represent; most often they are just
indicating that one character is filled with the ink color when the
other just draws the outline using an empty pattern (so in that case
too, the representative glyph uses such a monochrome fill pattern).

We also have the case where some scripts are using emphasizing colors
as a distinctive feature of the script, by tradition, even if there is
also an alternative representation using other emphasizing graphic
features (such as overstrokes, underlines, or box enclosing), but here
again the color (part of its name, such as "red" opposed to "black")
will be the normal semantic of the character.

When the character is renderable with its implicit colors in such a
way that the rest of the presentationan features can preserve the
distinction, these normal colors should still be used. Otherwise the
alternative using alternate glyphs or fill patterns or additional
conventional strokes will be used.

As another example (this is just a theoretical example, not a proposal
for encoding ), if we want characters to encode country flags, a flag
for France and a flag for Italy would normally only preserve their
distinction with their usual colors with each character unambiguously
encoding ench country flag distinctly. The alternative using
monochromatic glyphs would require using patterns to make the
distinction between the two tricolor flags, or could use additional
letters "FR" and "IT" drawn on top of a monochromatic flag with
undinstinctable three vertical bands.

So effectively the difficulty is to choose the distinctive
representative glyph using the most common conventions when color is
not preserved. But nothing should limit the ability of encoding
characters with their normal distinctive semantic colors. As this
represetnative glyph is not mandatory but only informative, this
fallback representation in monochromatic charts is not really a
problem. And charts cannot show all the necessary distinctions between
lots of characters (just consider the various dots using in various
scripts for example, or letters A in Latin, Greek and Cyrillic). The
glyph is not enough to determine the identity, you have to look at
other distinctive propoerties (including non normative properties such
as transliteration schemes).
Received on Tue Jul 03 2012 - 19:31:12 CDT

This archive was generated by hypermail 2.2.0 : Tue Jul 03 2012 - 19:31:13 CDT