> In most of the units AnsiString is used. But with this kind of
> string a real UTF8-coding isn't possible.
> I have a string stored in a field of a database (Charset UTF8) and
> a IB_Connection with Charset UTF8. The content of this field are
> characters from russian und spanish language.
> In IBODataset.pas there is a UTF8Decode (Row 3627) from the read
> content as ansistring! Why you are not using unicodestrings?

You must remove (or comment out) all those Utf8Decode and Utf8Encode
calls. Then it will also work with connections with connection
character set UTF8. This bug was discussed a long time ago and there
never was an official fix for it. AFAIR there are not a lot of
occurencies of these calls.

Best Regards

Stefan