The Shape of My Attention

I have ADHD. I'm not going to turn that into a symptom list or a redemption story. What follows is a plain account of how it shows up in the way I build software, written in my own terms rather than a clinical one.

The clearest pattern is hyperfocus. When a problem interests me, I disappear into it. A debugging session that starts at ten in the morning can still be running at four, and by then I've missed lunch, ignored two calendar reminders, and forgotten about the smaller task I'd flagged that same morning as more urgent. Discipline hasn't failed in that moment. It's latched onto the wrong target and kept running long after it should have handed control back.

The opposite problem is, if anything, more stubborn. Renewing a certificate, bumping a dependency with no interesting bug attached, writing documentation nobody's asked for yet, these sit untouched for weeks. I can tell you each one takes fifteen minutes. Knowing that has never once made the fifteen minutes easier to start. The gap sits between understanding that a task is small and being able to begin it, and that gap is where most of my real friction lives.

New ideas pull hard too. A tool idea or a small side project shows up mid-task, and it's more interesting than whatever I was already doing, simply because it's new. I have folders full of half-built utilities that each solved a real problem for two or three weeks of my attention before stalling. A few I still use today. Most died right where the interesting part ended and only the boring part, packaging things up, writing a test, was left.

None of this responds to willpower for long. I spent years assuming the fix was trying harder, and that approach never survived an ordinary Tuesday. What has worked better looks more like environment design than self-control. If a task is broken down to something like open the file instead of fix the bug, I can start it without making a decision first, and that turns out to matter more than motivation ever did. A timer sitting on the desk helps too, mostly because I have no internal sense of elapsed time once I'm absorbed in something. The rest is mostly about keeping the next step visible and singular, since a full plan asks me to hold the whole shape of a project in my head, and that's exactly the kind of thing I lose track of first.

Building this way turned out to have a side effect I didn't expect. Small, visible, low-friction next steps are just good engineering practice, regardless of anyone's attention span. Commits that do one thing. A next action that names itself instead of a vague goal. I didn't pick these habits up because a book told me to. I picked them up because I needed scaffolding, and the scaffolding happened to be good scaffolding for anyone.

I don't think any of this makes me faster than someone without ADHD, and I don't think it makes me slower in any fixed way either. It makes my output uneven in a specific pattern: long stretches where nothing visible happens, followed by disproportionate progress on whatever just became interesting. Working with that pattern, instead of pretending it isn't there, is the only approach that has actually held up for me over time.