zrlw

maybe i miss something, my benchmark result show little performance optimization. https://github.com/zrlw/dubbo-hessian-lite/blob/writeOptimization/java-8-test/src/test/java/com/alibaba/com/caucho/hessian/io/GlobalCacheBenchmark.java

Actually, this PR is just for the convenience of contributors to fix major formatting errors. The result of the repair depends on the contributor‘s JDK version, It's difficult to achieve...

Unless ```spotless:apply``` is performed on github PR action, it is difficult to ensure that the ```spotless:apply``` results from contributors are consistent with GitHub ```spotless:check``` action. https://github.com/apache/dubbo/pull/15391 but it seemed that...

> Hi, I'm trying to improve support for Java `record` types, but since `record` is only available from JDK 16+, I'm not sure where to place the related unit tests...

we take a temporary solution: ``` ReferenceCache refCache = DubboBootstrap.getInstance().getCache(); ProviderB providerB = refCache.get(referenceConfig); // set check true temporarily for invoker available checking. referenceConfig.setCheck(true); try { // check invoker available...

the next release version (1.2.1) of maven-build-cache-extension will fix it.

Notice: the build-cache extension will be activated only when using Apache Maven from 3.9.x https://github.com/apache/maven-build-cache-extension/pull/350/files

in your machine it take how many times to process SerializeSecurityConfigurator addition with this pr? what's the different of cpu profile between the pr and original codes? could you provide...

If it is necessary to add sample testing, you'd better add a separate test module instead of updating dubbo-samples-async-simple-boot, and submit a [dubbo-samples](https://github.com/apache/dubbo-samples) pr for it.

@AlbumenJ PTAL