What I did is modified MySql config file in /etc/my.cnf

[mysql] default-character-set = utf8 I am using MySql version 4.1.8 and up.

To create database with utf8 support:

CREATE DATABASE `foo` DEFAULT CHARACTER SET utf8 COLLATE utf8_bin

When creating a table with utf8 support:

CREATE TABLE `bar` (...) DEFAULT CHARSET=utf8 COLLATE=utf8_bin

Next you will need to change your db connection string to include "charset=utf8;" .

You can watch the server log file (assuming you have --debug in your mysql start script in /etc/init.d/) for command "SET NAMES utf8" if you really want to make sure it's setting the client encoding properly.

What I can't figure out is why do some Japanese, Chinese, Arabic and Korean characters come out as ??, wrong encoding, when I set the table and db collation to utf8_general_ci instead of utf8_bin?

I can only get true utf-8 support to work with default collation set to utf_bin when working with TEXT & BLOB fields.

Cheers,

-Daniel.