On 7-1-2012 18:29, Vander Clock Stephane wrote:

>
>> I think you're talking about raw UTF-8 bytes; as other have
>> suggested, you should be using CHARACTER SET OCTETS. Which
>> means no characters, just bytes (octets).
>>
>
> yes sorry i m confuse about character, code point or raw UTF8 byte...
> actually i prefere to use ASCII instead of OCTECTS because ASCII feet
> better to store ASCII char (that what is RAW UTF8 char)

The Firebird character set ASCII allows (or should allow, not sure; see
the bug that was posted earlier in the thread) only upto 0x7F as ASCII
is only defined for 0 to 127 (0x00 to 0x7F), so you won't be able to
store higher order all bytes that are generated by UTF-8 encoding.


--
Mark Rotteveel