I use arch, btw
The title is a cliché, I know. A little wink, a little euphemeric humour, the sort of phrase that has become more meme than meaning. But it also happens to be true again: I recently moved from Windows back to Arch Linux. It’s been a few years since I last lived in that world, mostly because the Linux kernel simply didn’t play nicely with some of the hardware I was using at the time. That constraint has softened, the ground has shifted, and I found myself returning—not to perform the stereotype the title implies, not to announce some badge of purity, but to test something more fundamental: what it feels like when the operating system stops trying to be everything at once.
The most immediate difference wasn’t philosophical. It was visceral. Boot times, application start times, the sensation of waiting for the machine to become itself—dramatically improved. And what makes that harder to ignore is that the hardware story here is not a “brand new rig” story. The core of this system is roughly eight years old, with the usual slow evolution we all do when we refuse to throw away something that still works: the case has changed twice, the CPU cooler once, a TPM chip had to be replaced, I added more RAM a couple of years ago, and I did the single biggest quality-of-life upgrade of the modern era by moving from SATA to a PCIe NVMe SSD. But those changes don’t explain the sharpness of the contrast. What I felt was not the power of new metal, but the weight of new software—the difference between a system that tries to predict your needs by default, and one that asks you, again and again, what you actually want.
And of course, that kind of contrast doesn’t stay confined to the desktop. It leaks into everything you build. Switching back to Arch sent me down an old rabbit hole: the tendency to over-implement, to over-design, to bloat an interface beyond what is necessary simply because it is possible. Not because the user asked for it, not because it adds meaning, but because it scratches the developer itch of completion. It made me slow down and look at Valeon with a different lens. Not “what can I add,” but “what does adding this cost,” not only in time, but in complexity, maintenance, performance, and attention.
In that light, I’ve started arguing with myself that Valeon is already, in a quiet way, feature-complete. Which doesn’t mean frozen. It means the next phase is less about stacking new pillars and more about tightening bolts, smoothing edges, and letting the house breathe. So the announcements on that front, for now, are mostly quality-of-life: there’s now a functional Calendar page, RSS feeds have a dedicated page, word highlighting should be fully functional across posts—including those with LaTeX—and if anyone notices any drift, I’d genuinely appreciate a report via the Contact page, because the smallest misalignment is the kind of thing that hides in plain sight until someone else’s eyes find it. The audio player has also been revamped so it feels like part of the page rather than an embedded afterthought, with improved styling and a waveform scrubber powered by WaveSurfer.js instead of a simple range input. Under the hood, the audio pipeline has been refactored, brittle parts reinforced, and a few lingering bugs finally put to rest. Beyond that, the more serious question has been hovering over me: how much more actually improves the experience of reading and listening to a blog, and how much more is just motion that feels like progress? Going forward, I want to be more deliberate—less “add because we can,” more “add because it meaningfully serves the reader.”
The same pattern has been settling over GigFin too. As I mentioned in the last announcement, it has reached the shape I originally needed it to have, and now it feels relatively complete: expense tracking is in, data exports are in, the dashboard is improved and customisable, Vehicle Profiles exist and integrate with expenses, state persistence is in place, and the dashboard has gained more widgets for seeing your numbers in ways that actually change behaviour rather than merely decorate it. I’ll almost certainly keep evolving the dashboard as my own workflow evolves—more widgets will likely appear over time—but I’m trying to resist turning GigFin into a kitchen-sink product, because that was never the point. It was meant to be lightweight and portable, a tool I can reach for from any device to track income and expenses, to surface key metrics, and to help me understand the rhythm of financial flow well enough that budgeting stops being a vague intention and becomes a lived practice.
At the same time, GigFin is FOSS, and that changes the posture entirely. What begins as personal utility can become communal usefulness, not by inflating into a monolith, but by staying clear, coherent, and welcoming to contribution. So if you’re using it, or you want to, and you have suggestions—features you genuinely need, improvements that reduce friction, edge cases you keep bumping into—I’m open to hearing them, and open to considering what belongs. The aim isn’t maximalism. It’s clarity. Not more for the sake of more, but better for the sake of better.
Maybe that’s the real meaning hidden inside the meme. “I use arch, btw” is funny because it’s performative. But returning to Arch, for me, has been a reminder that good systems don’t shout. They don’t overwhelm. They get out of the way. And if Valeon and GigFin are going to grow, I want them to grow in that direction: toward simplicity that is earned, toward features that justify their existence, toward software that feels less like noise and more like a clean room where your attention can finally do what it came to do.
