Building My Own Corner of the Internet

astrocontentweb

Every few months I have the same thought: my personal information and content lives on platforms I do not own, formatted by rules I did not set, in front of an algorithm I cannot see. So this year I finally did something about it and built this site.

Why bother, when Instagram, LinkedIn, and Medium already exist

Social platforms are convenient, but convenience is not the same as ownership. A post on LinkedIn is one algorithm change away from reaching nobody. A Medium article sits behind a paywall I do not control. Neither gives me a URL I can point to in ten years and trust that it still looks the way I intended.

I still want the reach those platforms if I choose to share there, so this is not a boycott. It’s just a more intentional way of information sharing: publish here first, distribute everywhere else second.

What it needed to be

A few constraints shaped every decision:

  • No database, no admin panel, no server process to babysit at 2am.
  • Posts are markdown files. Fixing a typo is a pull request, not a login.
  • It runs on my own home server, not a platform that can raise prices or shut down.
  • No comments, no newsletter, nothing that needs daily moderation.

How it is built

The site is Astro, rendered to plain static HTML at build time. Pages that do not need interactivity do not ship any client-side JavaScript, which keeps things fast without much effort on my part.

The look borrows the shadcn design tokens (colors, radii, spacing) rather than starting from a blank canvas, then layers a bit of personality on top: a teal accent, a subtle network background, and a few details I plan to keep refining.

Content lives as markdown and MDX files under version control. Publishing a post is a normal commit. When a new post lands on the main branch, a small pipeline drafts the LinkedIn post and opens a dev.to draft automatically, but nothing goes out the door without me reading it first and pressing publish myself.

Analytics run through a self-hosted Umami instance instead of a third-party tracker, so visits stay between me and my own server.

What is next

The plan is simple: create when I have something worth sharing, keep the infrastructure boring enough that it never gets in the way, and let the site slowly fill up with whatever I have been deep in lately.