I have a self-hosted instance of GitLab that contains a large portion of the code I have ever written with Git as its version control system. Powering this page is a script that queries GitLab for every commit that I authored and aggregates statistics on them. There are some caveats to these statistics, but this is a fun little project with some interesting insights nonetheless.

• Since these statistics only cover code I write on my own personal time, my activity in a professional seting is excluded.
• There is third-party and generated code commited and included in these results. For example, I use glad to generate GL loaders and that accounts for about 20,000 lines of generated code.

It is pretty clear that I opt to use C/C++ for most of my personal work. PHP looks like my typical choice for website backends. I suspect these results are skewed by generated code and third-party framework code. Languages in "Other" include lanuages such as CSS that, while in use, are proportionally very minimal.

There are prominent hills and valleys throughout the week. Closer to the weekend, I am usually more inclined to work on my personal projects. My activity trends downward as the week progresses, most likely due to work or classes. Then it trends back up for the weekend, probably due to the lack of work or classes.

Mornings are clearly a no-go and my late nights are pretty visible. My personal productivity ramps up as the day progresses, likely correlating with the free time I would have due to work or classes. It is a fairly consistent increase throughout the day, but my activity tops out soon after midnight. I suspect this because it is around the time where I would go to bed on days preceding a day of work or class. On free nights, however, this keeps going until a dead stop around 6 or 7, when I finally call it quits during late night grinds.