- NO.
- 023
- DATE
- Updated 2026-08-06
- READ
- ~3 min
- KIND
- Reference
- STATUS
- Reviewed
Personal blogs worth studying
A reference list for studying information architecture, navigation, archives, interactions, and long-term content structure.
Before shaping a personal blog, it helps to study sites that have already been running for a while. The useful parts are often the quiet ones: archive density, topic indexes, search placement, post metadata, and the way a site stays readable over years.
Primary references
- Antfu - Posts: a useful reference for archive density, year grouping, and lightweight page backgrounds.
- Innei: a useful reference for Light / System / Dark theme selection, article footer details, interaction entry points, and future extensibility.
- Cassie Codes - Writing: a useful reference for writing lists, dark mode, and long-page rhythm. (This site originally planned to use load more for long listings; it moved to static pagination at
/blog/page/N/on 2026-07-24 — appended content with no URL of its own turned out to be a real problem.) - Stitch via MCP: a reference for connecting design exploration with implementation workflows later.
Added 2026-08-06: media libraries, online editing, paywalls
This batch was studied with specific questions in mind — where media gets stored, how online editing is built, and how paywalls collect money. Every observation below comes from the public page itself, recorded here so later decisions have a source.
- Abo's Blog · Highlights: a game-clip video library in a card grid, with a live capacity meter on the page (
0 MB / 10.00 GB) backed by Cloudflare R2's free tier, plus support for externally hosted video links that cost no storage at all. The combination is the lesson: host the small files, link out for the big ones, and put the capacity bar where readers can see it rather than hiding the fact that the quota is finite. Its admin entry point sits on that same public page behind a single "admin passcode" field. That is the cheapest thing that works, at the cost of one shared secret: no record of who changed what, and no way to revoke one person. This site takes a different route, for reasons below. - lvyovo wiki: article pages are a client-rendered SPA (
lvy-neko) that paints a loading state first. The part worth stopping for is the writing surface at/write/<slug>: cover art, image management, a compression tool, and Markdown import all on one screen — the most complete online editor in this batch, and its information density is worth borrowing. - Xiaohu · AI digest: a fully built paywall — free sections stay open in full, deep dives and research show the first quarter, and pricing runs yearly / monthly / single-article, plus gift codes and team seats. It runs on a conventional account system (Google sign-in or an emailed magic link, paid by Alipay / WeChat / card), not on HTTP 402. Worth recording because this site's x402 experiment keeps drawing the question of whether anyone else is using it: this one is not. It solves the same problem — charging for a single article — on a completely different leg.
To study later
Later notes can add a short evaluation for each site: which structures are worth borrowing, and which design choices do not fit this writing workflow.
Comments
Comments are powered by GitHub Discussions. Sign in with GitHub to comment. Open the matching Discussion