Bytecode

Bytecode is an ‘intermediate representation’ of code.

Instead of directly translating high-level code (like Python or Java) into machine code that the CPU can execute, the high-level code is first compiled into bytecode.

This bytecode is fairly low-level (it mostly consists of simple instructions that a virtual machine can understand), but it’s not specific to any particular hardware architecture.

How bytecode compilers+interpreters work

When you run a program written in a language that uses bytecode (like Java or Python), the following steps usually happen:

1. The high-level code is compiled into bytecode by a compiler.
2. The bytecode is then executed by a virtual machine (VM)

So you can see that actually, bytecode languages are kind of a hybrid between compiled and interpreted languages!

Advantages of bytecode

• Portability: you’ve probably heard of Java’s slogan ‘write once, run anywhere’. And even though it’s often altered to ‘write once, debug everywhere’… the original idea still stands - that Java’s bytecode can run on any platform that has a Java virtual machine (JVM) installed.
• Performance: bytecode can be optimized by the virtual machine at runtime, which can lead to better performance compared to interpreted languages.
• Security: it runs under a virtual machine, which might give a sandboxed environment that could prevent certainn malware attacks (or just prevent poorly written code from crashing your entire system).
• Easier debugging: bytecode can retain more information about the original source code (like variable names, line numbers, etc), which can make debugging easier than debugging a fully compiled binary (anyone who’s used gdb before will understand :sigh:)

Bytecode can be used if the code needs to run on multiple platforms, so needs to be platform-independent.

Disadvantages of bytecode

• It’s slower than native machine code, since it has to be interpreted or compiled at runtime by the virtual machine.
• It requires a virtual machine or interpreter to run, which adds an extra layer of complexity.
• All users need to install the correct virtual machine (interpreter) for the bytecode to run, which can lead to compatibility issues and general annoyance.
• It can often be much less efficient in terms of RAM usage (which in this day, isn’t ideal… $600 DDR5 anyone?)
• Bytecode can be reverse-engineered more easily than native machine code, so if a company wants to keep their code closed source, bytecode might not be the best choice.

Examples of bytecode languages

• Java
• Python
• C# and other .NET languages