Less, but better

Most frontend tooling talks seem to focus on adding things to your project. There's no problem that can't be solved by adding more bits to your build, right?!

But when I looked at our UI library, I realised all those bits *were* the problem. Solutions that had been standard had fallen into disrepair in just a few years. Fundamental upgrades were blocked and the developer experience was awful.

People were quick to propose new-and-shiny silver bullets to replace the old-and-busted silver bullets. But repeating the same actions seemed unlikely to yield a different result.

So I proposed a new strategy: use as little as possible, and treat complexity as a risk.

By making every dependency fight for its place, we gained an appreciation for why each piece was there. By considering the risk each piece added, we had a framework to decide which risks were worth taking.

This did not lead to a pointlessly ascetic solution, but it did result in a set of deliberate choices and some guiding principles to keep the build lean.

Perhaps this could be your next project as well: defined not by what you add, but by what you leave out.