From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Mon Jan 12 2009 - 03:34:39 CST
On 1/12/2009 12:21 AM, Christopher Fynn wrote:
> If the criteia for encoding compatibility characters is different from
> that of ordinary characters - then perhaps they should be encoded in a
> block seperate from ordinary characters and not mixed together or
> unified.
The problem with this type of thinking is that any set proposed for
compatibility contains with some probability characters that are in fact
"ordinary" characters, in other words it contains some characters that
could have been encoded on their own. A formal determination on a
character-by-character basis of which are true compatibility characters
and which are (or could be seen as) ordinary characters has not been
made in the past, so the precedent is one of mixed allocation. And even
if you wanted to make a firm distinction from now on, it's not always
easy to settle which is which.
> If this whole group is required for interoperability ~ then
> encode the whole group on one of the upper planes as a single block of
> emoji compatibility characters. This would give us the whole lot
> claimed necessary for interoperability and avoid the PUA or markup.
For interoperability you'd want to get full coverage, of course, that's
the goal. But characters that are related to existing characters are
sometimes better encoded in proximity to their counterparts, than
together with characters from the same compatibility set.
That's been the practice in Unicode from the beginning. Over the long
run, users are better off it such practices are followed more or less
consistently over time.
A./
PS: It's interesting, actually scary, how many suggestions are being
made in this discussion that would change rather well-established
practice in extending the Unicode standard.
This archive was generated by hypermail 2.1.5 : Mon Jan 12 2009 - 03:36:10 CST