Well this has already been an interesting experiment. Only one week in, and I already feel more motivated and focused to work on personal projects and the goals that I’d set. Some of that might just be the ephemeral “New System Energy” that is plentiful when you start a new organizational system or habit, but what the hey; even if it’s fleeting, I’ll gladly ride it as far as I can!

What I Did

Gitea Actions is up-and-running, and I learned a bit more about Helm and Docker along the way.

Last night I got inspired and whipped up a project that I’d been thinking about for months. Sadly it is clock-time-dependent so I can’t really test and confirm that it works for another few days, but still - for someone who’s had difficulty completing or remaining focused on personal projects, I can’t tell you how good it feels to have actually gotten something from “this doesn’t exist” to “as far as it’s possible to tell at this point, this works” over the course of a few hours. That’s a habit I’d like to keep cultivating!

On that topic - as I mentioned in the previous weeknotes, I had intended to experiment with AIDE/AI Dev Tools over this week, and…holy shit. I am embarassed to admit just how confidently-wrong I was¹ about their utility. I’ve only scratched the surface of Cursor - I’m sure there’s oodles more functionality I’m not aware of, as well as better “prompt engineering” skill I can develop - and even so, the experience was unsettlingly transformative.

Don’t get me wrong - the ethical question of the morality of capitalizing a remixing of human creative output still remains, as do environmental concerns and the fact that this is another labour-hostile tool in the capitalist’s arsenal. This is not an unalloyed good, by any stretch. But I can no longer in good conscience continue insisting that these tools are all hype and marketing. They might not be capable of all the wondrous feats that salespeople say, but they are absolutely a tool that a thoughtful competent engineer can use to magnify their capability manyfold. We live in interesting times, for sure.

What I’d Like To Do

• Plumb in secrets from Vault to a Gitea Actions workflow (an obvious use-case would be re-building my “Telegram message on a build pass/fail”, as well as the image publication step)
• I mentioned in the title that this could arguably be “month-notes”, as it’s the start of a new month. I don’t think I’m going to start keeping these notes on that scale, but it does give me a prompt to get back into my old GTD routine. What with one thing and another I’ve let it lapse in 2025, but the mark of a well-formed habit is one that you can restart when it lapses. A monthly review to figure out the higher-level projects I want to work towards, and to be mindful of the passage of time rather than burning the days, would be a good idea.
• Ported over from last week - getting Keycloak working again on my homelab
• “Closing the loop” on my backup systems and verification. As I’ve mentioned before, I have BackBlaze ostensibly backing up my data (including the source code on my Gitea instance), but I have tested it exactly zero times. A good habit to bake into a Monthly Review!
• Honestly, I am excited to crank out some more coding tasks (as opposed to homelab maintainance and setup, which tends to be more configuration-wrangling and installation than coding) and see what these AI tools can do. I remember reading a comment from a fan a few months back saying that they had “started running out of side project ideas to do because they were being completed so quickly” - at the time I scoffed, but now, I can see it…
  • First and foremost, though, should be “productionalization” of the mystery unnamed project from the “What I Did” section, and a blog about it.

• AI
• CI/CD
• Gitea
• Productivity
• Weeknotes