I've been learning Python for about two weeks and recently updated my MiniGameStore project to version 1.7. I learned about classes and methods, then tried using them to add a restore-purchases feature. Because the class was stored in a separate file, I ran into issues with variables, imports, and a JSON file, and the whole process became much more confusing than using a simple function would have been.
I understand the basic theory behind classes, but I'm not yet convinced they provide a meaningful advantage for this project. When did classes start to make sense to you? Is my Game class a reasonable use of object-oriented programming, or would a dictionary and a few functions be simpler and more appropriate? I'd also appreciate advice on organizing the classes into separate files, handling the JSON data, and understanding why global variables are generally discouraged. Feedback on the overall project and suggestions for a practical, real-world approach are welcome.
5 Answers
Classes often become more valuable when you work with a library or framework that expects you to define objects in a particular way. For example, some database and ORM tools let you describe records as classes and then provide searching, validation, and other behavior automatically. Using those APIs can make the purpose of classes much more obvious than a tiny beginner project does.
Keeping a class in a separate file is mainly an organization choice, not a requirement. A small project can keep related code together in one module. Move a class into its own file when the module is getting crowded, when the class has a clear independent responsibility, or when you expect to reuse it. Separate files do not automatically make the design better; clear boundaries do.
Try not to rely on global variables for important application state. Globals can be changed from anywhere, which makes it harder to tell where a value came from and makes testing more difficult. Passing data into functions or storing it in a well-defined object gives the program clearer dependencies. For JSON, it can help to have one small component responsible for loading and saving the file, then pass the resulting data to the rest of the program instead of letting every class access the file directly.
You don’t need classes for every project. Functional code using functions and data structures can be perfectly appropriate, especially for a small program. Classes become useful when you want to keep related data and behavior together, represent several similar objects, or make a part of the program easier to reuse and test. They can improve organization and readability, but they also add structure and complexity, so there’s no benefit in forcing them into places where functions are clearer.
Your Game class sounds like a reasonable use of a class. It represents one concept and groups the data and operations associated with it. A dataclass may be worth exploring because it can generate the initializer and other standard methods automatically when the class mostly stores fields rather than complex logic. That reduces boilerplate without requiring you to write everything by hand.
I still wonder whether a dictionary plus a function would be enough here. I understand that the class is valid, but I’m trying to identify a concrete advantage rather than use one just because it’s a feature I recently learned.
A useful rule is to introduce a class when an object has both meaningful state and operations that belong to that state. If it only stores a few values and has no real behavior, a dictionary, named tuple, or dataclass may be simpler.