The Humble Doorstop: On the First Time I Considered a Wedge as a Safeguard
It’s a simple thing, a doorstop. A rubberized wedge, a heavy metal shoe, perhaps just a chunk of wood. We use it without thought, kicking it into place to hold a door ajar for a breeze, to move furniture, or to let the cat in. It is an object of pure, uncomplicated utility. Until, that is, you watch it fail.
I was in a small server room, a converted closet really, wrestling with a balky switch. The door was heavy, fire-rated, and possessed of a hydraulic arm that was supposed to hold it open but had long since given up its ghost. I did what anyone would do: I grabbed the nearest object that wasn't a server—a dense, rubber doorstop—and wedged it firmly under the door. It held. The air from the hallway flowed in, a minor victory against the stifling heat of the machines. My focus narrowed to the blinking lights of the network panel.
Then came the shudder. A heavy cart rolled down the hallway outside, its vibration traveling through the floor. The door, imperceptibly at first, began to shift. The wedge, which had seemed so secure, surrendered its purchase with a quiet, almost apologetic scrape. The door swung shut with a definitive, thunderous click. The lock engaged. I was sealed in a silent, humming tomb.
In that moment, the doorstop transformed from a simple tool into a profound lesson in operational philosophy. I had treated it as a permanent solution, a set-and-forget mechanism. But it wasn't. It was a temporary bypass of a designed safety system (the broken door arm), a single point of failure with no feedback loop. It offered no alarm, no log entry, no metric to indicate its impending capitulation. Its failure was binary and absolute.
We do this all the time in our digital domains. We prop open critical processes with what we assume are reliable wedges: a hastily written script that ‘always works,’ a manual backup routine we swear we’ll automate ‘next week,’ a log file we’re sure we’ll never need to read. We trust them implicitly because they have not yet failed us. They are our digital doorstops, and we never consider the vibration that will eventually dislodge them.
Now, when I look at any system, I look for the wedges. I ask: what is holding this open? Is it a designed, resilient mechanism with feedback and redundancy, or is it a simple, silent object that could be kicked away by the slightest tremor? The most reliable systems aren't those with the fewest points of failure; they are those that assume failure is inevitable and have a plan for the door swinging shut. They don't rely on the wedge. They fix the arm, or they install a second door, or they ensure someone is always there to hear the scrape before the click. The goal isn't to eliminate the humble doorstop, but to never be surprised by its simple, predictable nature.
Notes & further reading
A few pages I came back to while writing this: