Carson Gross on Dirty Code vs. Clean Code and Testing Strategies

Challenging "Clean Code" Ideology

• 💡 Carson Gross argues against the rigid adherence to "Clean Code" principles, particularly those popularized by Bob Martin.
• 🎯 He aims to reassure younger developers that alternative, successful coding styles exist and they shouldn't be intimidated by dominant ideologies.
• 🔑 Gross's approach, which he terms "quick and dirty," contrasts with conventional "clean code" recommendations.

Key Areas of Disagreement

• 🚫 Short Methods: Gross believes long methods are acceptable, especially if they represent important logic and refactoring doesn't offer clear code reuse.
• 🚫 TDD/Unit Testing: He prefers integration or end-to-end tests over strict Test-Driven Development and unit testing of individual functions.
• 🚫 Class Density: He advocates for denser, larger classes that might mix concerns, contrary to the single responsibility principle often emphasized in clean code.

The Case for Longer Methods

• 📏 Gross doesn't have a strict line count but feels antsy beyond 100 lines. He prioritizes code reuse as the primary driver for extraction.
• 🧠 He believes important functions should be large and unimportant ones small, making it easier to grasp core logic without excessive navigation.
• ⚠️ Extracting functions incurs costs like naming commitments and potential constraints on future changes, which he feels are often underestimated.
• 📊 Empirical studies suggest that functions up to around 200 lines of code may have fewer bugs per line, challenging the aesthetic-based assertion that longer methods are inherently bad.

Testing Philosophy: Integration Over Unit

• 🔍 Gross prefers integration tests or end-to-end tests because they focus on higher-level truths and tend to be more resilient to refactoring.
• 🐌 He finds unit tests often test implementation details, leading to brittle tests that break easily during refactors, creating a burden in large codebases.
• 🚀 While acknowledging the speed and ease of writing unit tests, he believes their focus on low-level details can hinder progress.
• ⚠️ End-to-end tests, while valuable, can be fragile due to their distance from the code, requiring a careful balance to remain useful.

Development Trade-offs

• ⚖️ Gross emphasizes that all development tools and approaches have trade-offs, and wisdom lies in finding the "golden mean" rather than blindly following one methodology.
• 💡 He cautions against presenting any single approach as unilaterally the best, encouraging discussion of associated costs and benefits.