Shifting planning gears

Runt is a massive exercise for me. It's a challenge to design all the little code pieces into one coherent orthogonal sculpture. I think I might have learned something new just now.

A year ago when I was starting with Runt I had no clue how all the pieces I wanted in will work on their own let alone together. I was pretty sure that the UI will be the backbone, but was not sure what shape it should take.

Slowly I realized that I cannot build it part by part, but I have to go wide and work on multiple parts at once so I won't over- or under-engineer anything. Everything was subject-of-change. This worked well...

Partially.

And that's the problem.

I now have myriads of things with variable complexity, none of which is quite usable. Everything lacks polish and little quality-of-life things. When you add feature at one spot, other missing features are that much pronounced.

The more work I do, the longer the to-do list gets.

And I no longer get my motivation from how awesome the entire thing is.

Thinking about it for a week, I now realized it's the planning that has to change. It's so embarrassingly obvious:

I need to focus on each part separately:

  • I shall never be distracted by lack of polish in parts outside of my current goal.
  • I shall not change my goal until the part adds value to my daily work.
  • I shall get pleasure from sharing how awesome it is.

So there.