I'm testing for OdbcFb_v2_0_145_Win32.

The VARCHAR/CHAR columns are working fine with UTF8 connection.
But BLOB text column is not transliteration.
I tried change to the CP943C connection. This BLOB's transliteration is fine.

I think that UTF8 connectioin result is odbc driver side bug since build 144.

My environments:

- WindowsXP sp2
- Firebird V2.1.1
- Excel 2003 japanese edition.

/* Create database */
CREATE DATABASE 'Localhost:TestDB'
USER 'sysdba'
PASSWORD 'password'
PAGE_SIZE 16384
DEFAULT CHARACTER SET UTF8;
SET SQL DIALECT 3;

/* Create table */
CREATE TABLE Table1 (
K INTEGER NOT NULL, B BLOB SUB_TYPE TEXT SEGMENT SIZE 80
, CONSTRAINT PK PRIMARY KEY (K));

-- My test code
Sub Fbodbctest()
    Dim con As ADODB.Connection
    Dim recordSet As ADODB.recordSet
    Dim outPutSheet As Worksheet
    Dim row As Long
    Dim col As Long
    
    Set con = New ADODB.Connection
    Set recordSet = New ADODB.recordSet
    Set outPutSheet = ThisWorkbook.Sheets("Sheet1") 'Output sheet
    outPutSheet.Cells.Clear
'
    con.Open "DRIVER=Firebird/InterBase(r) driver; SERVER=localhost; DATABASE=TestDB;" & _
        "Extended Properties=""UID=sysdba;PWD=password;CHARSET=UTF8;"""
' "Extended Properties=""UID=sysdba;PWD=password;CHARSET=CP943C;"""
  recordSet.Open "select * from Table1", con, adOpenForwardOnly, adLockReadOnly, -1
   
    'Field names to row
    For col = 0 To recordSet.Fields.Count - 1
        outPutSheet.Cells(1, col + 1).Value = recordSet.Fields(col).Name
    Next col

   'Field values to row
   row = 2
    While (Not recordSet.EOF) And row < 60000
        For col = 0 To recordSet.Fields.Count - 1
            outPutSheet.Cells(row, col + 1).Value = recordSet.Fields(col).Value
        Next col
        recordSet.MoveNext
        row = row + 1
    Wend
    
    recordSet.Close
    con.Close
End Sub

--
Regards,
Minoru