Hello All, DataSketches Memory 6.0.0 for Java 21 has been released.
*About DataSketches-Memory* This component of the DataSketches library provides high-performance access to off-heap memory for Java applications. These are low-level functions that are used by other DataSketches components. It does not provide any command-line interface or GUI. *NOTES:* - This Release of DataSketches Memory is for Java 21 and will compile and run on Java 21. It may not run on later versions of Java and will not run on earlier Java versions. - It uses the new java.base/java.lang.foreign (Panama, Foreign Function & Memory (FFM) API, [JEP 442](https://openjdk.org/jeps/442)) instead of relying on JVM internals. You may need to add the JVM flag `--enable-preview` at compile and run-time. - The FFM API has changed considerably since Java 17 thus the API changes in this release will reflect some of those changes. The primary change affecting this release is the use of the `Arena` class for allocating and closing off-heap memory. The `ResourceScope` class of Java 17 has been removed and its functionality simplified and moved to `MemorySegment.Scope`. - This release also fixes issues that were found in the prior release 5.0.0: - Please read the introduction to the datasketches-memory https://github.com/apache/datasketches-memory/pull/249. - A bug discovered by Paweł Wiejacha. See https://github.com/apache/datasketches-memory/pull/253. - A bug discovered just before we almost released RC2. It is documented in https://github.com/apache/datasketches-memory/pull/260. - A special "Thank You" to @frankgrimes97 for his considerable contributions to this release. *LINKS:* 1. DataSketches Download Site: https://datasketches.apache.org/docs/Community/Downloads.html 2. Release Summary: https://github.com/apache/datasketches-memory/releases 3. Developer build and test documentation: https://github.com/apache/datasketches-memory/blob/6.0.0/README.md 3. DataSketches Website https://datasketches.apache.org Thank you to the PMC members and community for taking the time to review this release! And a special "Thank You" to *@frankgrimes97* for his considerable contributions to this release. Lee. [email protected]
