The Case for “Impractical” CSS
CSS art tends to attract a predictable chorus of dismissals: “That’s not practical,” “Just use SVG.” It’s a flawed argument even before you factor in that no one is obligated to be practical at all times. But the debate misses a more interesting point: the techniques behind elaborate CSS art are often the same ones you’d use in a production stylesheet.
I was reminded of this while watching Lynn Fisher’s live stream of her single-div process. Like many, I assumed single-div artwork relied on dense, pixel-like box-shadow stacks. Her method, however, was far more mundane: a series of layered, sized, and positioned background gradients.

That’s the real revelation—not the technique itself, but the audacity of applying it. I’ve used gradients, layered images, and positioned backgrounds on client projects countless times. None of the CSS involved was new, complicated, or radical. The stream didn’t teach me anything about the properties; it changed my perception of what those simple tools could achieve.
Gradients in the Wild
Within weeks, that mental shift made its way into production code. The scenario was the perfect low-hanging fruit—a custom SVG felt slightly too bulky for the job, but a plain background wasn’t enough. A few layered gradients, paired with custom properties for easy tweaking, gave the effect without the overhead of an external asset.
Shortly after, the same approach landed in an official client component library, this time on Stacy Kvernmo’s suggestion. It worked cleanly and became a reusable pattern rather than a one-off hack.
The Practical Takeaway
The value of CSS artists isn’t just the pretty pictures. It’s the demonstration that the tools we use every day are more powerful than we give them credit for. The techniques aren’t esoteric—they’re standard properties applied with a bit more courage. Thanks to Lynn and the many other CSS artists for showing how much farther this language can be pushed, and for reminding us that our “Very Serious” tools are capable of things we haven’t tried yet.



