The old code converters are invoked with the following test program even
there are NIO code converters (as extended charsets) are available.

public class JConv {
    public static void main(String[] a) {
String msg = "\u3053\u3093\u306b\u3061\u306f\u4e16\u754c";
byte[] b = msg.getBytes();
String s = new String(b);
System.out.println(s);
    }
}

Its output on Solaris is:

% setenv LANG ja
% java -version
java version "1.4.2-beta"
Java(TM) 2 Runtime Environment, Standard Edition (build 1.4.2-beta-b02)
Java HotSpot(TM) Client VM (build 1.4.2-beta-b02, mixed mode)
% java -verbose JConv | grep EUC
[Loaded sun.io.ByteToCharEUC_JP from .../jre/lib/charsets.jar]
[Loaded sun.io.CharToByteEUC_JP from .../jre/lib/charsets.jar]
% setenv LANG zh
% java -verbose JConv | grep EUC
[Loaded sun.io.ByteToCharEUC_CN from .../jre/lib/charsets.jar]
[Loaded sun.io.CharToByteEUC_CN from .../jre/lib/charsets.jar]

However, both the old and NIO code converters are used in javac.

% javac -J-verbose JConv.java | grep EUC
[Loaded sun.io.ByteToCharEUC_JP from .../jre/lib/charsets.jar]
[Loaded sun.io.CharToByteEUC_JP from .../jre/lib/charsets.jar]
[Loaded sun.nio.cs.ext.EUC_JP from .../jre/lib/charsets.jar]
[Loaded sun.nio.cs.ext.EUC_JP$Encoder from .../jre/lib/charsets.jar]
[Loaded sun.nio.cs.ext.EUC_JP$Decoder from .../jre/lib/charsets.jar]

###@###.### 2002-10-10