Link to JEP: https://openjdk.java.net/jeps/400
Many standard Java APIs defer to the standard JDK charset, which itself is determined by the host's OS and locale. To get around this, many developers specify the charset when invoking these APIs. For instance, the FileWriter class constructor has an argument for specifying the charset. Even with these additions, many developers rely on the JDK's default charset, which itself is not standard across machines. A Windows machine with an english locale will have a different charset than a Windows machine with a Japanese locale. In JDK 18, This has been solved by making UTF-8 the default charset UTF-8 across all programs. To demonstrate this, look at the following code:
-
FileWriter fw = new FileWriter("test.txt"); // cp1252 in Java 17 and below (machine-dependent) -
FileWriter fw = new FileWriter("test.txt"); // UTF-8 in Java 18
Line number 1 was executed with a Java 15, line number 2 was executed with Java 18.
The following program prints out the default charset and native encoding for this JVM and machine respectively.
Prerequisites: Two JDKs, Java 18 and Java 17 or below.
Usage
[Java 18 (Windows) Assuming JAVA_HOME points to JDK 18]: java .\src\App.java
Then compare the output to running the program on versions below Java 18.
[Java 17 and below (Windows)]: 'C:\Program Files\Java\<JDK>\bin\java.exe' .\src\App.java