The "daily build" framing on this site started as a constraint experiment: ship something small, most days, across whatever discipline felt right that day, and let the project log be the receipt.
What it actually produced
Not forty polished case studies — closer to a dozen finished pieces and a long tail of smaller experiments that fed into them. The generative grid piece exists because of three failed noise-field sketches nobody will ever see. That's the actual value: the habit lowers the cost of the attempts that don't work.
Where it broke down
- Photo and video weeks compressed everything else. A single shoot day eats the "daily" budget for design or code that week — the cadence is realistically per-discipline, not literally per-day.
- Documentation lagged the work by weeks. This post exists roughly two months after the habit that prompted it started, which is its own small argument for writing things down closer to when they happen.
The goal was never forty finished things. It was never going eight days without making anything.
Adjustments going forward
Loosening "daily" to "most weeks, across at least two disciplines" — which is closer to what actually happened anyway — and starting the write-up the same day as the build instead of batching it later.