212 points roansh 1 hour ago 65 comments

Saw Karpathy's HackerTube and wanted the same thing for books. So I had Claude build one (Opus 5.5).

It's ~16.5k books that at least two different HN commenters brought up, from 2007 till now. Each book links back to the comments that mentioned it. Data is downloadable from the about page.

To be clear: Claude wrote almost all of it, including the data pipeline and the site. I mostly steered and made calls, bought the domain name, etc.

It counts mentions, so it's not perfect (praising a book and trashing it both count). If your favourite book is missing or matched to the wrong thing, tell me.

beanjuiceII 1 hour ago | parent

[flagged]

dang 1 hour ago | parent

We've banned this account for repeatedly (exclusively?) posting unsubstantive comments and ignoring our requests to stop.

If you don't want to be banned, you're welcome to email hn@ycombinator.com and give us reason to believe that you'll follow the rules in the future. They're here: https://news.ycombinator.com/newsguidelines.html.

roansh 1 hour ago | parent

Haha yeah. Looks like 226 different people brought it up, so it's in the top ~100. That's the "counts mentions" issue I guess. Also, the homepage shows random books (shuffled daily)

semilin 1 hour ago | parent

You'd have to be incredibly ignorant to suppose it to be a work not worth reading. It is not the best piece of writing, not even of Marx's, but its influence on humanity is enormous; that should suffice for the value in reading it.

blooalien 50 minutes ago | parent

Even if you completely disagree with every single word, it's still good to know what it actually says, if for no other reason than to be better equipped to argue against the parts you disagree with. Same thing for the Bible. It's one thing to know it backwards and forwards if you're a "true believer", but you certainly can't hope to hold any sorta worthwhile discussion on the topic with someone who is if you've never even read the thing. Knowledge is just another form of power; one that many willingly and actively choose to deny themselves for some reason.

cudla100 33 minutes ago | parent

I did read it and it still stole my time (how apt). Just read wiki.

Imustaskforhelp 1 hour ago | parent

I am really loving what people are building on similar ideas close to HackerTube and I am tinkering something of my own as well with these new ideas.

Awesome Idea and I would love to see what else community tinkers with things like these :-D

Hackernews is really a gold-mine in many ways.

Edit: I will try to read thinking fast and slow because of HN recommendations. I have heard some good things about the book but I will try to read it in future given HN's immense recommendations, so your website has already had a negative net positive for me :-D

blooalien 58 minutes ago | parent

> Hackernews is really a gold-mine in many ways.

It really truly is... I've learned so many new things, and discovered so many nifty softwares, books, and websites that I never would have found on my own, all thanks to HackerNews. Y'all folks here surely do have some really great tastes in "nerdy" stuff sometimes. ;)

LeComedien 1 hour ago | parent

Ayn Rand so high, though!

runtime_terror 1 hour ago | parent

Let us hope it was mentions about how idiotic of a book it is...

jmathai 1 hour ago | parent

I love this idea and hope the trend continues. It could even lead to finding niche communities to join.

jchook 1 hour ago | parent

The site is really fast. I see HTTP/3, OpenGraph, and Cloudflare. Can you expand on the technologies used?

Also... "Each book links back to the comments that mentioned it."

> Please don't put generated text in HN posts.

Claude speak gives me a gag reflex these days.

roansh 1 hour ago | parent

Thanks! First version was clunky, and hot-linked all cover images. Right now its plain JS and static JSON files, hosted on Cloudflare. First load pulls a ~30kb index of the top books, the full 16k list loads in the background, and everything has hashed filenames so it's cached forever. Covers are on R2. The data pipeline is python + duckdb over the HN dump.

And fair on the post, I had claude draft it and only edited it a couple of times. Should have just written it myself.

tonymet 28 minutes ago | parent

Local search index is the key to responsiveness.

vigilantpuma 1 hour ago | parent

The “mentions since” picker appears to be broken, nothing in the list changes as far as I can see. Also I wish it wouldn’t add another page to my history.

peri-cl 1 hour ago | parent

If this is the bug-reporting subthread: "Show more comments" stops working after clicking it once.

roansh 52 minutes ago | parent

It only keeps 30 comments per book so the second click was the last one :) pushed a minor fix, will look into adding all comments later

roansh 53 minutes ago | parent

Good catch, thanks! The filter was using a book's most recent mention on HN, which is pretty much always recent for popular books, thats why nothing changed? It now filters by when a book was first mentioned instead. and filters shouldn't add to history anymore.

vigilantpuma 11 minutes ago | parent

Much better now, thank you!

drayfield 1 hour ago | parent

I like the idea, but the wording might need to change as to how it counts a recommendation. As it stands Mein Kampf has been "recommended" by 651 HNers, a troubling statistic!

roansh 57 minutes ago | parent

Noted, pushing a fix for the wording for now. Will also look into improving the recommendations.

Edit: Changed it to "mentioned by"

danso 9 minutes ago | parent

If you’re using claude to build the whole thing and gather and parse the data, how on earth are you unable to get it to filter out the noise and show recommendations?

pseudony 18 minutes ago | parent

Hey, maybe the aspiring entrepreneurs just want to emulate the most successful tech bro’s ? ;)

sleiben 1 hour ago | parent

Thats pure gold. Thank you my friend!

RGS1811 1 hour ago | parent

This is wonderful. I'm delighted to see that Wittgenstein's Philosophical Investigations has been recommended over 100 times.

srik 1 hour ago | parent

I want this but for podcasts. I pick a podcast I subscribe to say the Tim Ferris show and I see book mentions by episode/theme/tim/guest with links to favorite/library/amazon. Life's too fragile for me to make something public like that but it's fairly easy is'nt it. Every passing year I'm surprised it doesn't exist yet, especially nowadays given how easy it is to obtain transcripts.

alexpadula 1 hour ago | parent

That would be cool

kxrm 55 minutes ago | parent

Build it!

Claude can turn something like this around in less than a day.

alexpadula 1 hour ago | parent

I expected way more books for software engineering.

phtrivier 1 hour ago | parent

The next stage is "HackerTime", a startup providing enough time to watch all the video and read all the books and play all the game and read all the news on "HackerZ for Z in tube, books, games, news".

With AI, in rust, invest before it's too late, etc...

polishdude20 1 hour ago | parent

I swear there used to be a hackernewsbooks website some years ago?

fnordpiglet 1 hour ago | parent

Slightly different take - use the Claude / OpenAI / muse desktop navigation integration to actually go into your YouTube account and like all the videos and watch them for a bit to seed your algorithm.

low_tech_punk 1 hour ago | parent

Next idea, build HackerNews, a website that features actual news from HackerNews.

jryle70 46 minutes ago | parent

Good idea. Maybe focus on posts that stay on the topic.

pndy 11 minutes ago | parent

https://news.ycombinator.com/from?site=hcker.news - filters out by keywords, domains and users

https://news.ycombinator.com/item?id=49308787 - keeps the default look while just removing AI-related submissions

I still have blind hope that tags feature with the obligation to include them in each submission will be added, but that would cut... "activity" some of the most prominent accounts.

alex_young 1 hour ago | parent

Don’t tell anyone, but I’m working on hacker hack

busymom0 1 hour ago | parent

Will that be a list of websites we should all hack?

alex_young 1 hour ago | parent

Don’t tell anyone, but I’m working on HackerHack, like HackerBooks and HackerBooks, but with comments and other posts and news and stuff. Some say it’s just HackerNews, but those are just Luddites who don’t get it.

bilater 1 hour ago | parent

Nice. I did something similar ranking the top 100 software/tool request of HN since 2008.

https://stillunbuilt.shiptiffin.app/

MYEUHD 55 minutes ago | parent

This was done already, back in 2021!

https://news.ycombinator.com/item?id=28595967

jjordan 4 minutes ago | parent

Funny, I bookmark interesting things all the time on here. I had this bookmarked. I've never revisited it.

amiune 47 minutes ago | parent

HackerMovies someone with free tokens please

QuantumNomad_ 42 minutes ago | parent

Well that one is easy!

- The Matrix (1999) https://m.imdb.com/title/tt0133093/

- Hackers (1995) https://m.imdb.com/title/tt0113243/

- Sneakers (1992) https://m.imdb.com/title/tt0105435

- Jurassic Park (1993) https://m.imdb.com/title/tt0107290/

- The Terminator (1984) https://m.imdb.com/title/tt0088247

- WarGames (1983) https://m.imdb.com/title/tt0086567/

There you go, that’s the entire list :)

amiune 34 minutes ago | parent

"Never send a human to do a machine's job"

drfuchs 2 minutes ago | parent

Don't forget the most prescient one!

- Colossus: The Forbin Project (1970) https://news.ycombinator.com/item?id=48142115

hyperman1 46 minutes ago | parent

I don't know if this is in there, but I wonder if books have a kind of 'eb and flood', where books raise to the top, stay there a while, then go away again. Could you plot some graphs? I'd expect some alrounders to stay around in the background forever, and other books come and go according to the average HN trending mood

roansh 39 minutes ago | parent

Love this idea, building a trends page now.

Built it - https://hackerbooks.org/#/trends

hyperman1 14 minutes ago | parent

I wonder if you've built it faster than I wrote that post.

mattfrommars 41 minutes ago | parent

I need to understand how do you guys build such slick or decent UI for websites. I get the backend but the front end, I have yet to crack. I know there is impeccable repo but I'm unsure how to go from here

radiospiel 37 minutes ago | parent

Using a coding agent really gives you that with minimal guidance. Try in opus session: "I want to build a website which looks similar to https://hackerbooks.org, but instead of showing books extract the top rated topics, render each toipic as a post it card" should give you a usable and maybe even good layout.

Starlevel004 29 minutes ago | parent

This is the stock claude UI

DuncanCoffee 39 minutes ago | parent

Add a way to suggest the correct author, sometimes just reading the first comment linked in the page has the right one, for example:

https://hackerbooks.org/#/book/OL999456W

> HIS DESIRE WAS FIERCER THAN A ROCKY MOUNTAIN STORM... HERS WAS RECKLESS AND DANGEROUSLY WILD

roansh 30 minutes ago | parent

fixing it now. a "suggest a correction" button is a good idea too

Erlangen 34 minutes ago | parent

I like the responsiveness of the website.

But there's no way Vanity Fair by Thackeray so high on "top all-time" list. It turns out that a lot of comments are about Vanity Fair magazine.

roansh 29 minutes ago | parent

yep, you're right, fixing it so it only counts comments that actually mention thackeray

5555watch 34 minutes ago | parent

After clicking an entry, below, "What Hacker News says", we see comments with mentions and surrounding text. It's now squashed, maybe you can add a function to expand on click, to see the comment with newline separation if such exist. Cause very often these comments presents a list of their favorite book, and an explanation of why. When these are squashed, it's not clear which explanation is preceding or following the entry.

Also, clicking the link "On "...."" it doesn't bring to the exact comment, only that thread or topic, so you need to search for it. It could bring directly to that comment.

pndy 27 minutes ago | parent

Not a fan of "For you" based on supposed similarities to what I've clicked before. Especially when e.g. "Scratch Programming Playground" has nothing to do with "Brave New World". I'd remove it but that's me.

Maybe language filters for non-English titles? Or if a book is in public domain already then links to the sources could be presented?

addandsubtract 18 minutes ago | parent

Have AI create a clean room rewrite of each book, so we can read them on the site.

laxisOp 16 minutes ago | parent

Gold gold . Thanks .now i need one for yt and one for only science papers

Izmaki 14 minutes ago | parent

oidar 15 minutes ago | parent

I'd love to see a books and youtube thing for metafilter too. It has a similarly good recommendation history for me.

yreg 15 minutes ago | parent

Hacker Music tomorrow?