Crosscut

A few months back, when I started working on the current iteration of Caterpillar, big things got done all the time. And that kept going for a bit, until recently. And now I'm getting lost in details. Bugs in the game, leading to bugs in the debugger, leading to rethinking tail call elimination, leading to revamping how the call stack works.

Part of it is, I've not been as laser-focused recently as I was before. But it's also natural. Early on, you make big steps. Day 1 you create something from nothing. Day 2 you still double what's there. Day 3 you only grow what you have by 50%. And this doesn't even consider that you don't just write new code. What you already have gets in the way, and regularly needs to be updated.

This ties in with the necessary switch from exploration to engineering that took me by surprise recently. So much of what I'm doing these days, is making the existing infrastructure more solid. And a lot of good is coming out ot it. I'm laying foundations, and it's going to make future work easier. I guess I just didn't expect to be doing so much of that yet.