Building This Blog

The implementation notes behind this static-first writing system.

This blog starts with a deliberately small surface area: posts are local MDX files, routes are statically generated, and metadata is derived from typed frontmatter. The goal is to keep publishing boring in the best possible way.

Why static-first

A personal blog does not need a database by default. Local content keeps the editing flow close to the code, gives every article a reviewable history, and makes builds predictable.

The content pipeline validates each post before it reaches the page layer:

type PostFrontmatter = {
  title: string;
  description?: string;
  date: string;
  category: string;
  tags: string[];
};

That means archive pages, tag pages, RSS, sitemap output, and Open Graph metadata all share the same source of truth.

What can grow later

The current version intentionally avoids accounts, comments, and a database. Those can be added when there is a real workflow that needs them. Until then, the site stays portable and cheap to operate.