Sequi Quo Quis Sit Earum Distinctio Odit Laudantium.: Best Practices That Actually Hold Up

The software industry generates opinions faster than it can validate them. In this post I want to cut through the noise and share the practices around Sequi Quo Quis Sit Earum Distinctio Odit Laudantium. that have genuinely held up across dozens of production systems and many years of real-world engineering work.

I made most of these mistakes myself before arriving at these conclusions. Learning from someone else's hard-won scars is one of the fastest ways to level up as a practitioner.

The One Principle That Unifies All the Rest

Before listing specific practices it helps to understand the single underlying principle that connects them all: optimise for the reader of the code, not the writer. Code is read far more often than it is written. Every decision should make the next developer's life measurably easier — and that next developer is frequently you, six months from now, with no memory of why you made a particular choice under time pressure.

"Any fool can write code that a computer can understand. Good programmers write code that humans can understand." — Martin Fowler

Practices Worth Genuinely Internalising

1. Favour Explicitness Over Cleverness

Clever one-liners feel deeply satisfying to write but are genuinely painful to debug under pressure. Prefer explicit, readable code even if it requires a few extra lines. The time you save writing the clever version is nothing compared to the time your entire team spends deciphering it later during an incident.

2. Keep Every Function Small and Focused on One Thing

A well-designed function does one thing and does it well. If you find yourself writing a function that clearly does two separate things, split it into two functions with precise, descriptive names. The ideal function length is five to fifteen lines. Anything consistently longer is a clear signal to refactor before moving on to the next feature.

3. Name Everything With Precision and Intention

Variable and function names are the primary form of documentation in any codebase. A well-named variable makes a comment entirely redundant. Actively avoid abbreviations and generic placeholder names like $data, $temp, or $result — they force every reader to carry unnecessary context in their working memory.

4. Make Invalid State Impossible to Represent in Code

Use typed value objects, enums, and strict constructor validation to make it structurally impossible to create objects in an invalid state. This approach eliminates an entire class of runtime bugs at compile time, which is always the better place to catch them from a cost perspective.

5. Test Observable Behaviour, Not Internal Implementation

Tests tightly coupled to internal implementation details break on every refactor, even when the observable external behaviour is completely unchanged. Write tests that describe what the system does from the outside perspective, not how it achieves that internally. Tests should function as your safety net, not as your ball and chain.

Patterns to Actively and Deliberately Avoid

• Deeply nested conditional blocks — use early returns and guard clauses to flatten the nesting and make the happy path obvious
• Functions with more than three or four parameters — introduce a dedicated DTO or value object to carry the related data as a coherent unit
• Boolean flag parameters in function signatures — they are almost always a sign the function should be two separate, clearly named functions
• Comments that explain what the code does line by line — the code itself should communicate that clearly; reserve comments for explaining why a non-obvious decision was made
• Magic numbers and bare string literals scattered through the code — extract every such value to a named constant with a clear description of its purpose and origin

How to Measure Whether Your Practices Are Actually Improving Things

1. Time to fully onboard a new developer decreases measurably from one quarter to the next
2. The bug rate per feature shipped trends downward when plotted over six-month windows
3. Engineers describe the codebase as genuinely pleasant to work in during retrospectives rather than something they dread touching
4. Feature lead time — from initial idea to code running in production — shrinks steadily as the codebase becomes easier and safer to extend

Final Thoughts

Best practices are useful heuristics, not sacred commandments. Understand deeply why each practice exists and you will develop the engineering judgement to know confidently when to apply it and when a pragmatic exception is genuinely warranted. The goal is always software that solves real problems reliably for real users — not software that perfectly adheres to a checklist written by someone who had never encountered your specific constraints.

What practices have you found most valuable in your own production work? Share your perspective in the comments below and let us compare notes honestly.