Combining Symfony with Inertia.js means writing controllers largely the same classic, Twig-like way, except that at the end a React component name with props is returned instead of HTML, making the entire frontend feel like a single page application, without a separate REST or GraphQL API ever coming into existence.

Many teams treat Kubernetes as the obvious choice as soon as more than one server is involved, yet very few applications actually justify the operational effort of a full cluster. Knowing the real signals for genuine need lets you make the decision based on facts instead of the latest conference talk.

EXTEND

DI

Deep Dive · Magento 2 Extensibility

Preferences vs. Plugins:
Which Extension Strategy When?

How interceptors work internally, why Preferences are usually the wrong choice, and the exact rules for Before, After, and Around plugins.

15 min read

Magento 2.4.8 · PHP 8.4

Magento Extension Points

Contents

1. 1. Two ways to extend Magento
2. 2. Preferences: replacing a class via di.xml
3. 3. Why Preferences are problematic
4. 4. Plugins (interceptors): the Interceptor Pattern
5. 5. Before plugin: modifying arguments
6. 6. After plugin: modifying the return value
7. 7. Around plugin: full control (and risks)
8. 8. Plugin ordering and conflicts
9. 9. Plugin limits: what plugins can't do
10. 10. Decision matrix: Preference vs. Plugin vs. Observer

One of the most common questions in Magento code reviews: "Should I use a Preference or a Plugin?" The answer is almost always Plugin, but why, and which type? This deep dive shows the exact differences, the internal mechanisms, and

Image signing proves that a container image actually originates from your own pipeline and has not been altered since the build. A software bill of materials lists every single component inside the image and makes supply chain risks visible before they turn into a security incident. Cosign and Syft from the Sigstore ecosystem implement both without requiring your own PKI infrastructure.

In JavaScript, the original error gets lost on re-throw: an Error is replaced by another one, and the context disappears. The error.cause feature from ES2022 solves exactly this problem. It allows structured error chains where each layer keeps its own context while the root error is passed through.

A covering index contains every column a query needs, so the server never has to touch the actual table. Anyone who applies this technique deliberately turns expensive random I/O against the table into pure, sequential index-only scans, often saving more time than any additional hardware would.

A TypeScript library that extends React, Vue or another host library must not bundle its types itself, otherwise two copies of the same type declaration collide for the consumer. Declaring peer dependencies in a type-safe way means setting up the right version range, the right imports and the right CI test matrix for exactly that problem.

GraphQL has no HTTP equivalent for business errors. Anyone who models business errors as resolver exceptions, fails to distinguish transport errors from network problems, and ignores extensions.category ends up with error handling that forces the frontend to react to heuristics instead of structure.

Snapshot isolation rightly enjoys a reputation as a comfortable compromise: it prevents dirty reads, non-repeatable reads, and most phantom reads, without the locking overhead of a full serializable level. Even so, it still allows a subtle but dangerous anomaly known as write skew, where two transactions independently make decisions that each look consistent on their own but together violate a business rule. Serializable snapshot isolation, or SSI, closes exactly that last gap. This article explains both isolation levels in detail and walks through a classic write skew example step by step.

An error inside an async function disappears without a trace, the console shows an unhandled rejection, an empty catch block hides exactly the problem you were looking for. Promise error handling follows its own rules, different from synchronous try catch. This article shows the concrete pitfalls and how to reliably avoid them.