Scoped Values enters incubation phase
Record Patterns enters preview 2
Pattern Matching for switch enters preview 4
Foreign Function & Memory API Enters Preview 2
Virtual Threads enters preview phase 2
Structured Concurrency enters the second incubation stage
Vector API (Vector API) enters the 5th incubation stage
Tiezi, JDK 20 / Java 20 has been officially released, and this version number is simply amazing.
JDK 20 is more than a simple upgrade. It is a release full of new features, bringing more choices for Java developers. With JDK 20, developers can more easily build complex applications, but also better manage and maintain these applications. This version not only includes some new features, but also improves and optimizes existing features.
JDK 20 is a short-term maintenance release and will be supported for six months. That said, it will be updated and supported for the next six months. This means that if you are using JDK 20, you can use it in production with confidence knowing that you will have adequate support and guarantees.
According to the Java development plan, the next LTS version will be released in September 2023, which is JDK 21. This release will be a major milestone for Java as it will be a long-term maintenance release that will be supported for up to eight years. This means that if you are looking for a stable and reliable version of Java, JDK 21 will be your best choice. 
This release includes 7 JEPs in various stages of incubation and preview. These JEPs, short for Java Enhancement Proposal, describe enhancements, new functions, and features of the Java SE platform and JDK. In addition, this release includes hundreds of minor enhancements and thousands of bug fixes covering all aspects of the Java SE platform and JDK. These enhancements and fixes provide Java developers with a better development experience and greater stability, making it easier for them to build high-quality applications 
:
Scoped Values enters incubation phase
In the Rust community, Scoped Values entered the incubation phase. They can share immutable data within and between threads, which avoids race issues that arise when using global variables. There are many benefits to using Scoped Values, such as their advantages over thread-local variables, especially when using a large number of virtual threads.
Record Patterns enters preview 2
Record Patterns are a new language feature that allow the deconstruction of record values. This allows us to manipulate and process complex data structures more conveniently. The combined use of Record patterns and Type patterns enables powerful, declarative, composable forms of data navigation and manipulation.
Pattern Matching for switch enters preview 4
The latest development of the Java programming language is that it has added a new pattern matching mechanism for expressions and statements, that is, pattern matching for switch expressions and statements, and an extension to the pattern language. This mechanism is very practical and allows us to express our intentions more concisely and safely when performing data-oriented queries.
Using pattern matching in switch allows us to test expressions against some patterns, allowing us to control the flow of the program more flexibly. In addition, this mechanism can also help us reduce code redundancy, because we can better reuse code.
To use this mechanism, we simply need to learn some new syntax and rules. In this way, we can more easily write efficient, safe and easy-to-maintain Java code.
Pattern Matching for switch enters preview 4
The latest version of Java introduces a new API called the Foreign Function & Memory API (Foreign Function & Memory API), which is now in its second preview phase. The purpose of this API is to enable Java programs to interoperate with code and data outside of the Java runtime. This means that, by efficiently calling external functions, and safely accessing external memory, Java programs can call native libraries and manipulate native data without the vulnerabilities and dangers of JNI.
This new API provides a more convenient and secure way for Java programmers to interact with native code and data. Now, Java programmers can more easily call native libraries without worrying about the vulnerabilities and dangers introduced by JNI. In addition, this API also provides some new functions, so that Java programmers can operate local memory more conveniently, so as to better handle local data. These new features include: pointer types, structures, unions, enumeration types, C function pointers, and more.
Structured Concurrency enters the second incubation stage
In JDK 19, an approach to multithreaded programming called structured concurrency was introduced. This approach aims to simplify multithreaded programming by treating multiple tasks running in different threads as a single unit of work by using the structured concurrency API. It should be noted that this approach is not intended to replace java.util.concurrent, but to optimize it.
One of the benefits of using structured concurrency is that error handling can be simplified. Since multiple tasks are treated as a single unit of work, errors can be caught and handled in a structured way. Another benefit is increased reliability. Structured concurrency can help avoid some common multithreaded programming mistakes, such as deadlocks and resource contention. At the same time, using structured concurrency can also enhance observability. Because multiple tasks are treated as a single unit of work, their progress can be tracked and monitored more easily.
Vector API (Vector API) enters the 5th incubation stage
Vector computing is a computation consisting of a series of operations on vectors. The vector API was born to express vector calculations, and by using it, we can compile at runtime to the optimal vector instructions on the supported CPU architecture, thereby achieving higher performance than equivalent scalar calculations. In addition, the goal of the vector API is to provide users with a concise, easy-to-use and platform-independent solution for expressing a wide range of vector calculations. Therefore, users can perform vector calculations through the vector API without worrying about compatibility and performance issues.