> A finished project is a dead project. If it isn't improving, it's dying.
This might be true for large and broadly scoped projects, but for small Unix-style modules/tools it isn't always the case.
To use real-world examples, see point-in-polygon[1] and robust-point-in-polygon[2]. The tools are both "finished" and any further API scope/modifications would be harmful to the software that has come to depend on them.
They could forever be improved in documentation, discovery (i.e. a website), etc but at a certain point it is better to place your effort into new tools.
This might be true for large and broadly scoped projects, but for small Unix-style modules/tools it isn't always the case.
To use real-world examples, see point-in-polygon[1] and robust-point-in-polygon[2]. The tools are both "finished" and any further API scope/modifications would be harmful to the software that has come to depend on them.
They could forever be improved in documentation, discovery (i.e. a website), etc but at a certain point it is better to place your effort into new tools.
[1] https://github.com/substack/point-in-polygon
[2] https://github.com/mikolalysenko/robust-point-in-polygon