That Randy Brown

Twenty Six Weeks of Code

When I first started writing, a year or so ago, I spent most of my writing cycles writing about changes to how my notes pipeline worked. I did this until notes eventually went private, as they are now, and then automated. Once the publishing of a note became divorced from the writing of a note, I stopped writing about it.

And indeed, For some time, I hadn't been doing anything public. When I was running anvil, all but the most basic utilities lived in private repos on Gitlab. However, for the last twenty-six weeks, I've written almost all of my personal code in public--even code for Happy Acorn Ventures.

This hasn't had the effect in one way I'd expect it to, given the cycle of producing to produce. I haven't written much, or at all, about what I'm working on, what I'm learning, why I'm doing it. In some ways, I wonder if this isn't because of the backlog of notes existing; I sometimes wish older ones would publish, and seeing a backlog of notes that, if I wrote one today, it would maybe not publish for a year, is sometimes disheartening.

The last half year of code have built a lot of muscle, though. I started with only one commit a day--a familiar pattern from back when I first started learning to program. Over time, as I got more comfortable with each milestone, I'd add more to the norm. Today, I write four commits a day. It takes me roughly an hour in the evening after the kids are in bed.

Perhaps, aside from the granularities of each individual project, which I'll likely touch on another time, the most important thing I have learned--or relearned--is the habit of being able to prime the next day's work, to make snowballs roll across days. Perhaps the hardest days have been those where I've reached a milestone on the majority of my projects and have to start without momentum on something again.