Archive: January 2025

1. Docker and Xdebug: A Clean Setup Without the Pain

   Jan 31, 2025

   Xdebug only works reliably in Docker containers once you understand the three most common pitfalls: incorrect host detection, missing path mapping, and Xdebug overhead during normal operation. Solve these three issues properly once, and debugging in Docker becomes just as comfortable as debugging locally.

   Read more

2. Tr: Practical Everyday Use Cases in the Shell

   Jan 31, 2025

   tr is often seen as the simplest text tool in the shell and gets underrated because of it. Character classes, targeted deletion with -d and squeeze with -s make tr the fastest tool for input normalization, delimiter swapping and cleaning up control characters, often quicker to type and quicker to run than an equivalent sed command.

   Read more

3. Marketplace Liability Under Section 25e VAT Act: What Online Sellers Must Know

   Jan 31, 2025

   If you sell through Amazon or eBay in addition to your own online shop, sooner or later you will run into a rule that sounds very technical at first glance but is quite simple in practice: the marketplace needs proof that you are correctly registered for tax purposes. Without this proof, your account can be blocked, with no wrongdoing on your part at all. We explain what this is about and what you can calmly take care of.

   Read more

4. Skeleton Loading States in Hyva Theme: Improving Perceived Performance

   Jan 31, 2025

   A skeleton screen replaces an empty white area with a rough preview of the coming content, making the wait feel shorter than it actually is. In a Hyva context with its GraphQL-heavy content loading, that is a powerful but easily misused tool that needs to be clearly distinguished from a classic spinner.

   Read more

5. Modal Stack: Managing Multiple Stacked Dialogs

   Jan 31, 2025

   As soon as an application needs to show more than one modal at a time, for example a confirmation dialog above an already open detail modal, a simple boolean flag stops being enough. This article shows how a global Alpine store manages a real stack of open dialogs, which z-index and focus rules apply per level, and how focus correctly returns to the triggering element when the top modal closes.

   Read more

6. Container Query Units: Using cqw, cqh, cqi and cqb in Practice

   Jan 31, 2025

   Container query units scale font sizes and spacing relative to the width or height of the surrounding container, not relative to the browser window. That means a card component automatically behaves differently in a narrow sidebar than the same card in a wide main area, without extra classes or JavaScript.

   Read more

7. WeakMap and WeakReference in PHP: Building Caches Without Memory Leaks

   Jan 30, 2025

   An ordinary array used as a cache with objects as keys or values prevents those objects from ever being collected by the garbage collector as long as the cache itself exists, and that is exactly what leads to creeping memory leaks in long running processes. WeakMap solves this problem structurally by removing the cache's own influence on the lifetime of the referenced objects.

   Read more

8. Setting Up the Hyvä Watcher and Hot Reload in Local Docker Development

   Jan 30, 2025

   A working hot reload loop decides in daily Hyvä work whether a layout change becomes visible within seconds or only after a manual deploy cycle. The Hyvä watcher runs in the background of the Docker container, watches phtml, CSS and JS files through a bind mount, and automatically triggers a Tailwind rebuild plus regeneration of the view_preprocessed directory on every change, without setup:static-content:deploy. Anyone who understands the mechanics behind it, sets the correct inotify limits and can check the watcher status cleanly saves hours every day compared to the classic production build cycle.

   Read more

9. Query DSL: Understanding the Bool Query in Detail

   Jan 30, 2025

   The Bool Query is the central tool of the Elasticsearch Query DSL for combining multiple conditions logically. Anyone who understands must, should, must_not, and filter, and knows when scoring context applies versus filter context, builds search queries that are both precise and fast, instead of throwing all clauses into one bucket and hoping for good results.

   Read more

10. Recognizing AI Hallucinations in Code

    Jan 30, 2025

    AI assistants like Claude occasionally produce code that looks plausible but relies on invented functions, classes, or packages that never existed. Unlike errors in prose, these hallucinations in code either fail loudly right away or quietly misbehave without any warning. This article shows concrete detection patterns and verification habits to catch them before such code reaches production.

    Read more