Decision: Jekyll Slug Migration

Date: 2026-03-23
Status: Complete

Problem

The blog was migrated from Jekyll. Post frontmatter slugs retained a .html suffix (e.g. my-post.html) from the Jekyll era. TanStack Router expects clean slugs without extensions.

What was done

The .html suffix is stripped at data generation time — when posts.ts loads the frontmatter, it removes the .html before the slug is used in routing or navigation.

getPostBySlug() includes defensive normalization: it tries both the raw slug and the .html-stripped version. This handles any edge case where old links or data might pass in a suffixed slug.

Scope

All 297 posts had their slugs normalized. No post URLs broke — the clean slug matches the new routes, and the .html suffix in old links would 404 at the Cloudflare Pages level (expected behavior; no redirect rules were added since there was no existing production traffic on the old domain).

Why not strip in frontmatter directly

Stripping at generation time rather than editing 297 .md frontmatter files keeps the source files clean and avoids a massive diff. The blog post content files in apps/blog/_posts/** are out of scope for duyetbot anyway — they belong to Duyet Le.