How this site works
This site is deliberately boring underneath, and that’s the point.
Every page starts as a plain-text Markdown file on my computer. A static site generator called Astro stitches those files into ordinary HTML, and the result is served from a plain file host. There is no database, no server-side code, no subscription, and no platform that could hold the content hostage. The whole thing costs about the price of two coffees per year — the domain — and everything else is free.
The design constraints I set for it:
- Own everything. Content is plain text in a git repository. If any provider in the chain disappears tomorrow, the site moves in an afternoon and no link breaks.
- Cheap to maintain. Publishing something new means writing a text file and pushing it. No admin panels, no updates to install.
- Modular. New sections, galleries, or interactive experiments can be added as new folders and components without restructuring what exists.
That last point is the part I care about most: the plan is for this site to eventually run small projects — sketches, simulations, toys — right in the page, not just describe them.
This post doubles as the first test of the publishing pipeline. If you can read it, the pipeline works.