Show HN: HackerBooks – HackerTube, but for books mentioned on HN

Original Article ↗ Hacker News Discussion ↗

A new site called HackerBooks aggregates and ranks books mentioned on Hacker News, linking each title to the original comments that referenced it. Commenters praise its speed, simple architecture, and clever use of local search and static assets, while also probing its reliance on AI for UI and data parsing, the pitfalls of treating all mentions as recommendations (e.g., controversial or sarcastically cited works), and metadata issues like misclassification and missing long‑tail titles. The project is seen as part of a broader wave of lightweight, AI-assisted curation tools built on HN data, sparking ideas for similar services for videos, podcasts, and other media, as well as reflection on how often books are being mentioned on the site over time.

Overall reception

  • Many commenters like the concept: a fast, HN-driven directory of books mentioned in discussions.
  • People see it as part of a broader trend (HackerTube, HackerBooks, spin‑offs for tools, Reddit, Wikipedia, etc.).
  • Several say HN is a “gold mine” for discovering books, tools, and other nerdy media.
  • Some are skeptical of treating HN as a taste oracle, noting ideological skew and demographic bias (e.g., heavily male, certain genres underrepresented).

Implementation, performance, and data sources

  • Site is praised for extreme speed: tiny HTML/CSS/JS, local search indexes, preloaded top‑books index, lazy loading of full catalog, edge caching via a CDN, HTTP/3.
  • Stack: static JSON + vanilla JS frontend; data pipeline uses Python and DuckDB over public HN dumps / BigQuery / API.
  • Local search index and per‑book JSON enable instant navigation and prefetching of comments.
  • Some minor bugs reported (filters not updating, history pollution, “show more comments” limit) and quickly patched.

Mentions vs recommendations, sentiment, and bias

  • Initial wording treated any mention as a “recommendation,” leading to absurd results for controversial or infamous books; changed to “mentioned by.”
  • Several argue strongly for distinguishing recommendations from neutral or critical mentions, handling sarcasm, and weighting by factors like karma or unique users.
  • Requests for clearer “most mentioned” vs “most recommended” views and more robust sentiment analysis.
  • Concerns about classification accuracy (e.g., magazine vs book confusion, misfiled genres, wrong publication dates) and missing long‑tail titles.

Feature ideas and UX feedback

  • Popular requests:
    • More filters (fiction/nonfiction, genre, date, language, format, page count, license, public‑domain / free links, 2020s, etc.).
    • Semantic search and better recommendation sections (“For you” skepticism).
    • Ability to surface obscure books (mentioned once or twice) instead of only popular ones.
    • Correction / feedback tools for titles, authors, categories, and covers.
    • Direct linking to the exact HN comment; expandable comment text.
    • Links to more sources (indie bookstores, global Amazon, library catalogs, free/DRM‑free editions).

AI usage and meta‑discussion

  • Creator openly used an AI assistant for the frontend and launch copy; some applaud this, others dislike the “AI voice.”
  • Commenters note that AI agents make small, static, plain‑JS sites newly attractive and easy to build.
  • Multiple spin‑off ideas appear (for podcasts, movies, music, dev tools, “HackerMedia,” etc.), many explicitly framed as “just ask an AI and ship.”

Longevity, prior art, and trends

  • Several similar “HN books” sites existed and died (404/abandoned); static hosting is suggested to improve longevity.
  • One commenter provides aggregate stats showing HN book mentions rising until 2022 then declining, prompting debate about reading habits, social media, LLMs, and changing user demographics.