Chapter 1: The Basics of Programming Languages

When discussing programming languages, two primary categories emerge based on execution methods: compiled languages and interpreted languages. However, I propose a third category: intermediate languages, which are compiled for a virtual machine, exemplified by Java's bytecode designed for the JVM. In practice, Java's bytecode is often converted to machine code via a Just-In-Time (JIT) compiler, complicating its classification.

Compiled languages employ a compiler to transform source code into specific machine code. Conversely, interpreted languages utilize an interpreter that reads and executes code on the fly. This distinction means that compiled languages require a complete compilation process before execution, while interpreted languages, like Python, allow for real-time modifications during runtime.

To create a compiler, developers must write software that generates executables. This task involves several components, including the scanner, parser, and various analysis stages such as syntactic and semantic. Additionally, an abstract code generator, optimizer, and machine code generator are typically required. The linker, which manages code dependencies, adds another layer of complexity. Consequently, the pool of engineers capable of developing compilers is relatively small.

In contrast, creating an interpreter can be simpler since it does not necessitate binary code generation. Generally, interpreters are compiled programs, although they can also be written in interpreted languages, resulting in a layered approach.

It's theoretically possible to write a compiler or interpreter in any language. For example, you can find resources online to create a C compiler using Java. However, certain languages like Java and Python are less suitable for this purpose. Historically, many compilers have been written in C or C++ due to their practicality at the time. Languages such as OCaml, Rust, and Scala are now preferred as they incorporate functional programming paradigms, providing syntactic sugar and features that enhance reliability during compiler development.

When selecting a language for compiler development, it's essential to consider the target platform for the resulting programs. Cross-compilation, or Canadian compilation, can also be conducted, particularly when developing new platforms. For instance, the reference compiler for the Java virtual machine is predominantly written in C++.

Rust serves as another example; its initial compiler was developed in OCaml, but later iterations were rewritten in Rust itself, utilizing the original compiler for self-porting.

Designing a programming language is a complex endeavor. While there are tools available online that facilitate the creation of simple languages with minimal coding, the results may lack sophistication.

This inquiry leads to a philosophical question: which came first, the machine or the software? I lean towards the notion that software precedes hardware, as machines require programs to execute algorithms efficiently. Without software, machines hold little value. However, this is a classic chicken-or-egg dilemma, reminiscent of discussions found in the history of mathematics.

Stackademic 🎓

Thank you for reading through to the end. Before you leave, please consider showing your support by clapping and following the author! 👏 Follow us on X | LinkedIn | YouTube | Discord. Explore more content at In Plain English | CoFeed | Venture | Cubed. Visit Stackademic.com for additional resources.

Chapter 2: Compiling vs. Interpreting

The first video, "The Programming Language Guide," provides an in-depth overview of different programming languages and their characteristics, helping viewers grasp the foundational concepts behind language design.

The second video, "How To Build A Programming Language From Scratch," walks through the process of creating a programming language, offering practical insights and techniques for aspiring developers.