Laracon US 2025

Defining Resilience in Modern Code Most developers focus on confidence—the belief that their code works under ideal conditions. However, Zuzana Kunckova argues that resilience is the far more critical metric. Resilience is the ability of an application to withstand stress, recover from setbacks, and adapt to failures without crashing. Whether it's a viral traffic spike or a third-party API going offline, resilient code continues to function. It doesn't just survive; it manages failure gracefully to protect the user experience. Defensive Input Strategies Never trust the user, the device, or the network. Common sense is rarely common in the wild. You must validate and sanitize every input early in the lifecycle. Beyond just preventing malicious attacks, clear validation ensures the system doesn't enter an unpredictable state because a user entered data you didn't expect. Catching these errors at the perimeter prevents 'garbage in, garbage out' scenarios from poisoning your database or crashing downstream logic. Proactive Error Management and Logging Exceptions shouldn't bubble up until they kill the process. Resilient systems anticipate potential failures and deal with them proactively. This means moving away from ugly default error pages and toward meaningful feedback. Simultaneously, your logging must be structured. Avoid 'hello' or 'test' logs. If a failure occurs in production, you need logs that provide actionable debugging insights, not noise. Alerts should notify you of silent failures long before a frustrated user reaches out to support. Implementing Graceful Degradation If one part of your system breaks, don't let it take down the whole house. If JavaScript fails to load or an image server is down, your HTML structure should remain readable and functional. Use cached content as a fallback when live data is unavailable. Whether it is a failed database connection or a flaky internet provider, retrying connections and serving 'stale' but useful data is always better than showing a blank screen or a spinner that never stops. Testing the Unhappy Path Testing only the 'happy path' is a recipe for disaster. Users will inevitably find ways to interact with your app that you never intended. To build resilience, you must test edge cases, service failures, and unexpected states. Stop thinking like the developer who knows how it should work and start thinking like a user who is trying to get things done on a spotty 3G connection. If your code can survive their chaos, it's truly resilient.