Two Scoops Study - 4

Chapter 6 of Two Scoops of Django focused on how to properly handle Model in Django. It was emphasized that models should be designed with sufficient thought, and if they grow too large, splitting into separate apps should be considered.

The methods of model inheritance were also organized together.

• Abstract base class
• Multitable inheritance
• Proxy model

The pros and cons of each approach were compared, along with discussion of when to use which type of inheritance. Alongside this, the study also covered widely-used patterns like TimeStampedModel, precautions during migrations, criteria for using Null and Blank, use of Binary field, and the pros and cons of Generic Foreign Key and generic relations.

On the design side, the study broadly examined normalization, caching and denormalization, PostgreSQL-specific fields, the _meta API, custom model manager, and the meaning of the default manager. In particular, it addressed the problem of the Fat model, and compared Model Behaviors and Mixin as ways to address it, as well as the role of stateless helper functions.

In the deeper study, database normalization was covered. The process of structuring data to reduce redundancy in relational DB design was explained, and this was connected to and understood alongside actual modeling intuition. Conversely, NoSQL was approached with a focus on Denormalization, data consistency, scale-out, and differences in the modeling process.

Finally, group modeling work was done for the book management service project. After listing the required fields, the normalization process was carried out, and table relationships were organized while considering actual query patterns and constraints. Through the study, the group reaffirmed the importance of deepening understanding to the point of being able to explain concepts—not just knowing them—and of the power of communicating theoretically with peers.