My Entropic Toolbox

When my Dad got home from work, I followed him down into the garage, where he nursed an aged British sports car back and forth through levels of disrepair.  I was too short to hold the light, so he got down the heavy cigar box full of nuts and bolts for me to sort. There were wood screws, machine screws, screws with deformed threads, brass, steel, zinc, like a tossed salad with a dressing of oil and dirt.  I never completed the task, and the next time I followed my Dad into the garage, the box was there, unsorted, as surely as the car was still broken. Some years ago I became a small time landlord, and therefore, a handyman. I got excited about well organized tools.  The hardware store was a big, beautiful library of solutions, and I bought a toolbox with lots of compartments and tools to go in it, but when I went to fix something, I often didn’t have what I needed.  An extra pair of pliers. A half-inch washer. A wooden shim.  The organic, analog problems I encountered had infinite variation, but my shiny new toolbox had discrete, limited categories.  I found myself looking around on the ground for a piece of wood or a pen cap or something shaped like a “J”.  Anything but another trip to Home Depot.

I was in grad school and was too busy to obsess over my tools, so my toolbox got messy.  When I installed a lock or fixed a disposal, some of the extra screws and fittings stayed in the box.  My toolbox, which once had perfectly sorted trays, became a salad of flotsam from previous jobs.  At first it was depressing that, not only was I incapable of bringing everything I needed to a job, but also had a messy toolbox, but as the mess increased and grew more diverse, it became a kind of primordial ooze out of which new ideas evolved. Now when Plan A fails, I root around in there, letting the objects run across the hole in my mind to see if one falls into place, like looking into my Dad’s cigar box of undifferentiated debris with a soft focus and letting the match present itself.  Old ideas have magnetic properties, and sometimes all you have to do is wave them around to feel them pull at compatible problems.

When I designed my first pattern library for HP in 2006, I learned this lesson all over again.  A major development effort yielded a robust, precisely defined set of components that were perfectly matched solutions to the problems we thought we had, but when it came time to build something with them, there was always some new requirement that demanded an exception. We had to break and rebuild allot of pieces before the toolkit was messy enough to be useful.

In my garage is an empty yogurt pint containing wood screws, machine screws, nuts, bolts, and other debris from my toolbox.  When my 4 year old daughter follows me in there, I tell her to separate the nuts from the screws, and that seems to occupy her long enough for me to fix something.