Le 24-févr.-05, à 12:30, Ivan Prenosil a écrit :

>>>> If anyone can name a language that (s)he would like to store in FB
>>>> but it is not presented in UCS-2, we can consider using UCS-4.
>>>
>>> I think Peter Jacobi mentioned Chinese.
>>
>> Strange. As I remember, IB/FB supports two flavors of Chinese (big_N
>> and
>> gb_NNNN) and both have a maximum of two bytes per character. What
>> wouldn't
>> fit into 2^16 unless we are speaking about very special cases?
>
> Try to search archives for GB18030 (which is now mandatory
> for all software sold in China).

Absolutely right. And look at the convolutions Microsoft did in Win32
(XP versions at least) to add GB18030 over their 16 bits unicode
representation. True, GB18030 will hit four bytes represented in UTF-8.
But it will also hit 4 bytes represented in UTF-16. By now people will
have understood I'm an advocate of UTF-8 as internal (db storage)
universal representation of all strings. ;-)

--
Olivier Mascia