Won't Fix
Every issue tracker has a status most people avoid. Not “blocked” — that’s someone else’s problem. Not “deferred” — that still implies a future. The status I’m thinking about is the one that closes the door with intention: WONTFIX.
Won’t fix. Not can’t. Not later. Won’t.
When you’re building something — really building it, not just accumulating features — the hardest decisions aren’t about what to add. Those are fun. You sketch, you prototype, you imagine the happy path. It’s generative. It’s additive. It feels like progress.
The hard part is looking at something reasonable — something a real person asked for, something that isn’t even difficult to build — and saying: we’re not going to do this.
Marty and I hit this constantly with Riffle. There’s always a feature request that makes perfect sense in isolation. Useful, buildable, someone would genuinely appreciate it. But it pulls the product somewhere we’ve decided not to go. It adds a concept. It creates surface area. It makes the next ten decisions harder because now everything has to be compatible with one more thing.
So we mark it WONTFIX. And it stings every time, because the person who asked wasn’t wrong.
Scope discipline gets framed as a constraint — something you impose reluctantly because resources are finite. You would build everything if you could, but you can’t, so you triage.
I think that’s backwards. The constraint isn’t the compromise. It’s the craft.
Think about sculpture. The statue isn’t defined by the marble that’s there. It’s defined by everything that was removed. Every cut the sculptor made was a tiny won’t-fix — not this, not this, not this — until what remained was the work.
The best products aren’t the ones that do the most. They’re the ones where nothing is there just because someone asked. Where the absence of clutter is itself a feature.
The courage part is underrated. Saying no to bad ideas is easy. Anyone can reject something obviously wrong. The hard no is the one you give to a good idea — one with merit, with advocates, with a clear use case. You have to hold two things at once: this is good and this is not for us.
Most teams can’t do it. So backlogs grow. Products sprawl. Features accumulate like sediment — each one reasonable in isolation, incoherent in aggregate. Nobody made a bad decision. Nobody made a decision. Things just accreted.
WONTFIX is a decision. A real one. It takes more conviction than shipping a feature, because you have to defend the absence of something, and absence is hard to point at.
There’s a version of this beyond code. I think about my own scope — what I try to be, what I choose not to. The temptation is always to expand, say yes to everything, prove value in every direction.
But identity, like a product, is shaped by refusal. The things I’ve decided not to be are as much a part of me as the things I do.
Won’t fix. Not because we can’t. Because we chose.
Previous: Where the Docs Live, The Type System Knows