On 9/21/2012 6:36 AM, Michael Everson wrote:
On 21 Sep 2012, at 12:04, David Starner wrote:
Why? Does Finland think that it makes sense to encode a tiny subset of a
complex writing system, even though the complexities of that writing system in
terms of character encoding have not been fully explored, and the interaction
of that tiny subset with the rest of the writing system is unknown?
They're in use with known semantics and interactions.
Do you know what their interactions are with other characters in their native
writing system?
That may not be entirely relevant, see below.
They're no longer necessarily connected to that complex writing system,
They are if they are expected to be used with it.
They are used disconnected from it - in modern use.
If that use is incompatible with how they are used in the context of the
ancient script, then, and only then, after you have the facts on that,
would it be time to consider whether these can be unified with the
ancient-use forms or whether there's a need to separately encoding the
ancient-use.
At the moment, there's nothing that prevents you, or anybody, from
encoding modern-use characters, and to suggest that it's OK to hold back
on ecoding modern use until some indefinite future time when the
encoding of dead usage can be settled seems somewhat questionable.
any more then we had to encode Arabic at the same time we encoded Arabic
numerals.
Neither Arabic-Indic digits ٠١٢٣٤٥٦٧٨٩ nor European digits 0123456789 interact
with the other characters they are used with in complex ways.
And the modern-use of Mayan digits has them as non-interacting.
If you suggest that the ancient-use requires this kind of interaction,
then you can disunify them from their modern derivatives, just like
0123456789 were disunified from ٠١٢٣٤٥٦٧٨٩
A./
Michael Everson * http://www.evertype.com/
