From: Michael Everson (everson@evertype.com)
Date: Wed Dec 25 2002 - 14:23:04 EST
At 12:01 -0800 2002-12-24, Asmus Freytag wrote:
>One of the problems is that modern computer character sets have been
>Latin+Cyrillic, allowing users easy access to Latin Q and W when
>processing Cyrillic. Essentially, existing character sets before
>Unicode, in other words legacy character sets, have set the
>precedent for this unification.
Easily said. But Asmus, you can't point to any Kurdish Cyrillic data, can you?.
>Disunifying this belatedly in Unicode would introduce non-negligible
>data conversion problems. This situation is exacerbated by the fact
>that Latin and Cyrillic Q and W do not look noticably different, if
>at all, which means that the even the (inadvertant) future use of
>Latin Q and W cannot be ruled out, perpetuating the incidences of
>dual encoding.
Same can be said of Aa and Oo. We've done this one before. SOFT SIGN
has a Latin equivalent. Cyrillic Q sometimes looks like a big q not a
Q. It's messy. Mixing alphabets is less good than encoding the
naturalized borrowed letters. In the view of some.
>Coptic does not have this legacy issue, for one there aren't any
>parts of 8859 that can be used for Coptic.
Yes it does. Anyone who has tried to use Unicode for Coptic has had
to use Greek. Jost Gippert for instance. (Who has said he will be
delighted to reencode all his data.)
>PS: the situation for Georgian is yet again different, in my view it
>has few analogies with either the Greek/Coptic or Kurdish examples,
>but I have nothing new to add to the discussion at this time so I
>elided the mention of it in the quoted text above.
The Georgian has to do with a false unification (whatever the reasons
for it, it was a mistake).
-- Michael Everson * * Everson Typography * * http://www.evertype.com
This archive was generated by hypermail 2.1.5 : Wed Dec 25 2002 - 15:15:36 EST