You Can Read the Source

Internet speed keeps getting faster every year, and today’s mobile devices often outperform older desktops. But just a few years ago, catering to mobile users was a delicate art. If you wanted to offer a smooth experience, you had to minimize the number and size of JavaScript files. A large file could freeze or even crash a mobile device. Desktop users on older browsers like Internet Explorer faced similar challenges.

To address this, many websites adopted the "lighter version" approach, redirecting low-performing devices to subdomains like m.example.com. Here, every single line of code was scrutinized for efficiency. You couldn't casually throw in a 2MB JavaScript file; you had to ensure every script was optimized and did only what was absolutely necessary.

This environment forced developers like my team and me to deeply understand the libraries we included. We weren’t just blindly copying and pasting code; we read through the source to figure out exactly how it worked so we could replicate only the functionality we needed.

For one project, I was tasked with removing jQuery. At the time, jQuery wasn’t considered large by modern standards, but it was the heaviest script on our page. While jQuery made programming accessible by smoothing over browser quirks, it was possible to replace many of its features with native code. We swapped $() with document.getElementBy* and replaced utility functions one by one. The trickiest part was detecting when the DOM was ready, a process that wasn’t standardized across browsers back then. Reading through jQuery’s source code taught me how it handled these quirks, and with that knowledge, we could implement a lightweight alternative.

This experience gave me a newfound appreciation for the underlying mechanics of the browser. jQuery was powerful, but we rarely needed everything it offered. Similarly, we replaced moment.js, another popular library at the time, with just a single line of code because our use case was so specific. By analyzing what we actually needed and reading the source, we saved time, bandwidth, and headaches.

This habit of reading source code extended beyond libraries to frameworks. I dove into Symfony and Ruby on Rails, both of which are beautifully structured and packed with lessons for anyone willing to explore. I even looked at a tiny framework called Hyperapp. Its simplicity was eye-opening, and I learned how to extract just the pieces I needed for my own projects.

More recently, I’ve been exploring React. Despite using it for years, its structure still feels unintuitive to me. However, digging into its source code has provided insights that even the documentation can’t teach. The patterns and trade-offs the React team made reveal valuable lessons about building scalable, efficient tools.

Reading the source isn't just about removing dependencies or slimming down your codebase; it’s about leveling up as a developer. When you take the time to understand how a library or framework works under the hood, you gain a deeper knowledge of the tools you’re using and the problems they’re solving. That knowledge, in turn, helps you write better, more efficient code.