I wouldn’t say bad necessarily, but I do operate under the assumption that I modeled it wrong the first time. And the second time probably too. It’s more of the idea of not waiting to have all the information, as the exercise of creating the model brings out information.

Trying the model out in practice for a week or month helps show where the weak points are, what invalid assumptions you had and what you missed. It’s more of an evolution.

But sometimes you have to wait for enough responsibilities to be floating around before you realize there is a concept there wanting to get out. I have found I’m much better at refactoring than designing (because my designs are usually wrong).

I used to feel a little bad about having a bad model after I realized it was bad. “Why couldn’t I see that earlier?” and such. But all I can do is model what I know, not try to model what I don’t know, and create an ecosystem that’s easy to change, so when I can see the right model, I have no fear of change.

What do you mean? This is legacy code, so no unit tests are currently in place.