Project Gutenberg is somehow still the best part of the internet

The first website I ever bookmarked, sometime around 2003, was Project Gutenberg. I was a teenager with a dial-up connection, a copy of Crime and Punishment I couldn’t afford, and a vague suspicion that the internet was full of things that adults didn’t want me to find. Turns out the most subversive thing on the internet was a website where some volunteers in Illinois had typed out the entire Western canon for free.

Twenty years later, JSeiko submitted gutenberg.org to Hacker News with a four-word title: “keeps getting better.” Within fifteen hours it was at 900 upvotes and 189 comments. That’s an absurd amount of attention for a website that has basically not changed its visual design since it figured out that beige worked.

The story in one sentence

Project Gutenberg is a 55-year-old volunteer project that types up out-of-copyright books, posts them as plain text and EPUB, and asks for nothing in return.

That’s it. There is no subscription. There are no algorithmic recommendations. There is no AI-generated summary. There is a search box, a list of the most-downloaded books this week, and 70,000+ titles you can read on a flight that takes off in twenty minutes.

Why this hit the front page

The article itself is just the front page of gutenberg.org. JSeiko didn’t write an essay. He linked the homepage. And HN flooded it because, in 2026, after a decade of every text-rich service on earth being enshittified into oblivion, the existence of something that just works and just gives you the thing feels miraculous.

The thread reads like a group therapy session for people who remember when the web was a public-library extension cord instead of a slot machine.

What the comments are arguing about

Three threads, roughly:

1. The donation tip. Several commenters point out - correctly - that the Project Gutenberg foundation runs on a tiny budget and a few hundred bucks of one-time donations is genuinely meaningful. The site has no tracking, no ads, and somehow still pays for its bandwidth.
2. The EPUB quality wars. The OCR has gotten significantly better in the last few years. Old scans of 19th-century novels used to be lousy with rn-instead-of-m errors; the new ones are clean enough that Kindle owners are casually sideloading them again.
3. The “but where do I start?” objection. New visitors get overwhelmed by 70k titles with no recommendations. The top reply, as always, is: download Pride and Prejudice and stop overthinking it.

Decision checklist

Read the original if…	Skip if…
You haven’t visited gutenberg.org in 5+ years and want to be quietly delighted	You’re hoping for new commentary - there is no article, it’s just the homepage
You want a list of the highest-quality EPUBs to sideload	You only read post-2000 books (PG is mostly pre-1929 by US copyright law)
You’re curious why HN keeps re-upvoting the same site every 18 months	You hate plain HTML enough that beige hurts your eyes

What I take away

There’s an unfashionable lesson in the existence of Project Gutenberg.

Almost every popular internet project of the last ten years was designed to be acquired, monetized, or scaled. PG was designed by Michael Hart, in 1971, to do one thing: take a book that nobody owns anymore and make it free to read on a computer. He typed the Declaration of Independence into a mainframe at the University of Illinois and called it volume one.

Fifty-five years and seventy thousand volumes later, it’s still doing the same thing. No pivot. No AI strategy. No B2B SaaS reinvention. Just a steadily growing list of books, maintained by people whose names you don’t know, given to you for free.

That’s the part HN keeps re-upvoting. Not the books. The proof that you can still build something good, slowly, in public, without an exit strategy.

---

Discussion on Hacker News · Source: gutenberg.org · Submitted by JSeiko