◀ Prev · Parts · Full · Next ▶

Originally, Rhino compiled all JavaScript code to Java bytecode in generated Java class files. This produced the best performance, often beating the C++ implementation of JavaScript run with just-in-time compilation (JIT), but suffered from two faults. First, compile time was long since generating bytecode and loading the generated classes was a resource-intensive process. Also, the implementation effectively leaked memory since most Java virtual machines (JVM) didn't collect unused classes or the strings that are interned as a result of loading a class file. (This changed in later Java versions.)

◀ Prev · Parts · Full · Next ▶