[LOG4J2-1636] Console Appender does not pick up Oracle Java 8's sun.stdout.encoding and sun.stderr.encoding

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 2.8, 2.8.1
• Component/s: Appenders, Layouts

Environment

Windows

• Assignee: Gary D. Gregory
• Reporter: Eldar Gabdullin
• Created: 09/Oct/16 17:27
• Updated: 30/Jan/22 16:18
• Resolved: 25/Feb/17 10:05

Description

The Oracle Java 8 platform supports two system properties to configure the System console: sun.stdout.encoding and sun.stderr.encoding.

---

Default log4j configuration prints to console using platform's default encoding, however that's not what standard windows console expects.

I.e.

System.out.println("Поехали"); // works just fine
System.out.write("Поехали!\n".getBytes(Charset.defaultCharset())); // characters are messed up
log.info("Поехали"); // messed up in the same way

This happens not just on one machine, but on all Windows desktops and servers in our organization.

Can we have out of the box Layout/Appender which prints directly to System.out/err?

Issue Links

• breaks: LOG4J2-1799 Error determining the current charset
• is related to: LOG4J2-783 PatternLayout's default charset should not be UTF-8
• is related to: LOG4J2-255 Multi-byte character strings are scrambled in log output
• is related to: LOG4J2-1038 Incorrect documentation for layout default charset
• relates to: LOG4J2-3372 ConsoleAppender on Windows does not use platform default encoding