Friction
Years ago I got pulled onto a project that was quietly falling apart. The interface had somewhere between fifty and sixty open UI defects. The component library sat unmaintained inside the main codebase, so nobody trusted it. QA had assumed someone upstream was checking the visual layer, so they had stopped looking at it. Nothing said what the design was actually supposed to be. I was embedded deep enough to feel all of it, and close enough to the exit to leave it for someone else.
I stayed. I pulled the design out into its own source of truth, something the whole team could point at and build against. The defects finally had somewhere to resolve to. The work I did to climb out of that mess outlived the project by years.
I didn't have language for why at the time. I have it now, and it comes from an odd place.
Friction only sharpens when one side holds still. Drag two identical hard surfaces across each other, the way two blades or two stones grind, and all you get is mutual wear. Both sides lose. Hold a stable, harder surface against a softer one, a whetstone against a blade, and the same friction that would have destroyed becomes the thing that sharpens. The asymmetry is the whole point. Without something steady to work against, friction is just damage.
That reframed a lot of my career. The thing I kept doing, on projects like that one, was not removing the friction. The fix was never to take the friction away, it was to put something solid in the middle for everyone to work against. When design and engineering grind with nothing stable between them, they wear each other down: rework, lost context, defects, the same argument every sprint. What they were missing was a shared reference, not a quieter room.
There's a second piece I only see in hindsight. The pressure of staying inside something hard, instead of backing out, is what produced the most valuable work I've made. I didn't plan the tooling that came out of that broken project. I endured the project, and the tooling was what the pressure left behind. Pressure, given enough time, makes the harder thing.
The image that stuck with me most, though, is a brake rotor. Leave one sitting and it rusts. The first time you use it again it squeaks, then it bites, then over a few cycles it smooths, and then it shines. Three things in that I keep coming back to. The shine was always under the rust, so friction reveals the surface, it doesn't create it. It takes repeated cycles, not one clean pass. And the rotor that never gets used is the one that rusts. The danger is disuse.
That last part is what I had backwards for a long time. I used to read friction as a sign something was wrong. Now I read the absence of it as the warning. The squeak is stage one, not failure. Most people back off right there and decide the new thing doesn't work, when the squeak is just the rust coming off.
So here is the pattern I can finally name. I run toward the friction other people route around, and I try to build the steady surface that makes it sharpen instead of grind. It has been the same move every time, from standing in the gap between design and code to the messiest projects I've taken on. I didn't feel like I was on a line while it was happening. Looking back, I was on it the whole time.
