“Enough” sounds like a decision made after the work: enough features, enough research, enough polish. I am trying to move it to the beginning. A useful constraint names what a project needs to protect before it names what the project could accumulate.

What is the work for?

For Lattice, enough means a stranger can find a note, read it comfortably, understand a connection, and return to the place they started. A graph is welcome if it helps that journey. It is not required to justify the app's existence.

For A neighborhood atlas, enough means one route is described with care and can be corrected by the people who use it. A large map with unexamined captions would be more complete geographically and less complete ethically.

Constraints as room

A page budget makes me choose the sentence that actually carries the argument. A ninety-minute experiment leaves energy for a second attempt. A monthly letter means I do not need every day's reading to become an announcement.

These are not universal ideals of smallness. Some work needs scale, redundancy, or painstaking detail. The point is to choose a size appropriate to the responsibility rather than inherit a preference for more.

The review question

Instead of “What else could we add?”, I ask “What is still difficult for the intended reader?” The answer might require a feature. It might require removing a feature, improving a heading, or admitting that the note does not yet know what it means.

Small bets & long horizons turns this into a rhythm: protect a long-lived question, bound the next attempt, and decide what evidence would justify another. Enough is not the enemy of ambition. It gives ambition somewhere to stand.