Red Hat has been a contributor to OpenJDK since its inception, and we are currently one of the most significant contributors to the OpenJDK project, both as developers of new versions and as maintainers of OpenJDK 8u and 11u. In addition to being a core developer of the Java Platform, Red Hat is also a heavy consumer of Java technology through our suite of Java-based Middleware and Application Services.
We welcome an analysis of current practices regarding the life cycle of Java releases to understand how Java consumers can be better served. Although the recent cadence changes announced by Oracle to move to a 2-year LTS cycle pertain solely to their proprietary JDK version, this will have an impact on OpenJDK distribution life cycles as well. This is because OpenJDK distributions have chosen to follow the same LTS cycle as the proprietary Oracle JDK to maintain consistency and reduce fragmentation of version usage in the industry.
As developers working on the Java Platform, we feel that the 2-year LTS change provides a quicker way for consumers of LTS versions to access newer Java features. At the same time, as developers of applications that are certified on LTS versions of the Java platform, we are concerned about the impact this will bring to Java application vendors like ourselves and our customers; certification of complex applications on a version of Java can often take many months and it has significant resource costs attached to it. The change from 3 to 2-year cycles means that ISVs will now have to either certify at a faster pace, or skip certain LTS versions and potentially fragment the ecosystem.
Major changes such as LTS cadence have potential implications across the entire Java ecosystem. We believe broader input into these decisions is essential to balance the needs of JVM developers, developers who code in Java, Java users, and Java application vendors. While we do not oppose changes to LTS cadence, we urge Oracle to seek broader input from vendors and customers who certify on the Java platform to determine if 2 years is indeed the optimal release cadence, or whether other options ought to be explored.