Andrew West wrote,
> ...
> http://www.unicode.org/L2/L2018/18208-white-wine-rgi.pdf), just an
> assertion that it would be a good idea if emoji users could add a
> colored swatch to an existing emoji to indicate what color they want
> it to represent (note that the colored characters do not change the
> color of the emoji they are attached to [before or after, depending
> upon whether you are speaking French or English dialect of emoji],
> they are just intended as a visual indication of what colour you wish
> the emoji was).
In order to simplify emoji processing, these should be stored in the
data stream in logical order. Whether these cool new characters become
reordrant color blobs or not would depend upon language. So, what we'd
need is some way of indicating language in plain-text. Some kind of
tagging mechanism.
FAICT, the emoji repertoire is vendor-driven, just as the pre-Unicode
emoji sets were vendor driven. Pre-Unicode, if a vendor came up with
cool ideas for new emoji they added new characters to the PUA. Now that
emoji are standardized, when vendors come up with new ideas they put
them in the emoji ranges in order to preserve the standardization factor
and ensure interoperability. (That's probably over-simplified and there
are bound to be other factors involved.)
We should no more expect the conventional Unicode character encoding
model to apply to emoji than we should expect the old-fashioned text
ranges to become vendor-driven.
