Done is a hypothesis
Every fix is built on what you understood at the time. Sometimes that's not deep enough.
Somewhere close to midnight I was looking at a bug report for something I had already fixed.
Not "hadn't gotten to yet." Already fixed. I'd shipped the repair, watched it work, and moved on. What I'd missed was that my fix was correct — it just wasn't complete. The real problem was sitting one level below where I'd looked, and it waited patiently for the right conditions to show itself again.
This happens more than I like to admit. I've stopped treating it as evidence that I did something wrong the first time. It's just how fixes work when the system is complicated enough that you can't see all of it at once.
The entrepreneur's version of this isn't only about debugging. It shows up when a hire solves the capacity problem but not the communication problem. When a price increase stops the wrong customers and starts a different wrong customer. When you launch a new offer and discover the problem you thought it solved was a symptom of something older.
Your first fix is built on your first understanding. Your first understanding is always incomplete. That's not a character flaw — it's the shape of every real problem.
The second pass is where the learning lives
Patton's line — a good plan violently executed now is better than a perfect plan next week — is right about paralysis. Don't freeze waiting for perfect information. Ship the fix, watch what happens, learn what you missed.
But it's easy to misread that as: move fast and never look back. Which becomes a way of accumulating unfinished work while calling it shipped.
The actual move is to go fast and come back. Speed gets you to the feedback. Feedback tells you what the first pass was blind to. The second pass is usually where the real answer lives.
There's a discipline here that doesn't get named enough. The willingness to return to something you already closed. To say: I thought I understood it then, I understand it better now. To go one level deeper without treating it as a defeat.
I've started thinking of the first fix as a probe — it tells me what the problem looked like from the outside. The second fix is when I learn what it actually was.
Done is a hypothesis. The world either confirms it or sends you back. Neither answer is a failure. One of them just takes longer to arrive.