A new client’s WooCommerce checkout was throwing fatal errors after a developer rebuilt field validation from scratch. The fix was not a better library but a short function hooked into WooCommerce, and here is why that was the right call.
A client wanted one checkout field, and the old code tempted me into a week-long refactor. I shipped the field with two WooCommerce hooks instead, a reminder that a working feature on time beats a perfect rewrite that never ships.
A WooCommerce client needed clean URLs for an old custom event plugin, and the Rewrite API alone kept getting fragile. Here is why a proper object-oriented routing system holds up better as a WordPress site grows into an application.
A WooCommerce add to cart button was failing silently for some users, with no logged errors to go on. Here is how a few lines in wp-config.php turn on WordPress error logging so you can see what is actually breaking on a live site.
A client’s custom events plugin had grown into tangled procedural code that broke every time they touched it. This is how object-oriented programming, and the pattern behind WordPress hooks, keeps a plugin maintainable as it grows.
A client added an email signup form with a heavy marketing plugin and wrecked his site speed. Here’s how to build a lightweight form that posts to your provider’s REST API instead, with full control and no bloat.
Duplicated pricing logic across half a dozen WooCommerce templates turned a 30-minute change into a day of risky edits. The DRY principle, and one function as the single source of truth, is the fix.
A junior dev fixed a 404 by flushing WordPress rewrite rules on every page load, which wrecked performance. Here’s how to register rules on init and flush them only once, when they actually change.
A WooCommerce checkout threw intermittent 500 errors and the logs were useless. Here’s how logging with context, the user, their cart, the gateway, turns a noisy debug.log into error logging you can actually act on.
A client had a WooCommerce checkout throwing random 500 errors while a 200MB debug log told them nothing. Here is the custom error logging setup I used to catch, count, and get alerted on new errors.