Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Meta book recommendations from Ask HN threads (mapfilterfold.com)
273 points by odie88 on March 27, 2020 | hide | past | favorite | 42 comments



MapFilterFold is a side project I’ve been kicking the can on for a while, it started out as a monstrous spreadsheet I maintained for a few months in 2018.

The site uses Amazon affiliate links.

There have been a few similar projects [1] [2] which create lists from links to Amazon and other stores. MFF differs by only using data from Ask HN threads about books and using book titles (or acronyms like GEB or SCIP) since most comments on those threads don’t include links. More about the data can be found on the About page.

This is also my first Elixir + Phoenix project, which was a joy to use once I got moving.

[1] https://hackernewsbooks.com/

[2] https://ramiro.org/vis/hn-most-linked-books/


Can you change the page titles to be more descriptive? I've bookmarked a few of these pages but had to manually change the titles to have the book title in them.


Hey! Sorry about that. I pushed out a quick update for individual book pages to be titled {MapFilterFold - Book: [bookTitle]}

Hopefully that makes most of the bookmarking easier. I'll work on a solution for other pages later today!


This is really cool, thanks for this! A lot of the books I read came from HN recommendations so this is pretty useful to me :)


On my wishlist is a book site that will also give you top positive and negative reviews for each book from high quality sources like Boston Review or The Atlantic. I like to pair reading a book with what's happened since or other ideas at the time.

I don't always remember to do this though so that's why getting these review recommendations from GoodReads (for example) would be ideal.

For example, parts of Thinking Fast and Slow have widely been discredited and the author has walked back claims. But you wouldn't know that from the book or its popularity.


After the Why We Sleep fiasco I'd quite like a site which peer reviews all the references in a book (could be crowd sourced wiki-style) individually. And then the book, or each chapter, gets a rating for factual correctness.


That is a good point. Also reminds me I'd like a site that lists all the references in a book so I can more easily search them on the computer and not necessarily need to even own the book.


This is how conventional journal publishing works, so I don't see why this couldn't be implemented for books which provide a bibliography.


That's a good thought. The next step is going in reverse, selecting the best referenced material and auto generating a quality non-fiction book from it. That's my dream.


Would Wikipedia not be the best place to do this (an objective evaluation of the literature)? With links deep linking into the book with Open Library (and Internet Archive initiative)?


Would you minding pointing to what the fiasco was about. I read the book based on all round positive reviews and recommendations.



I did too, and it turns out a lot of the claims are either made up (the literature doesn't support the claims, or doesn't seem to exist at all!), potentially harmful or misleading (modified plots to suit the narrative).

In hindsight some of the claims are just weird, like saying you can measure the effect of one bad night's sleep months down the line.

On the whole it got people talking about sleep, which isn't a bad thing. But, the peer review (see sibling posts) only looked at one chapter I think, so God knows what else is wrong.

I don't think anyone is doubting that you should probably get a bit more sleep, try not to blast blue lights in your face and get into a routine. Standard sleep hygiene stuff.



There have been a few discussions about it on HN as well, where Alexey Guyzey, author of the criticism linked, joins in on the conversation:

(4 months ago) https://news.ycombinator.com/item?id=21546850

(1 month ago) https://news.ycombinator.com/item?id=22419958



I've often wished for the same thing. I want to know what's wrong about the book before investing time in it.


Interesting that https://mapfilterfold.com/ and https://hackernewsbooks.com/top-books-on-hacker-news look to be attempting the same goal but with different results.

Curious what's different under the hood.


This is a great resource, thank you!

Tangentially related, it would be interesting to require book recommendations to always come with anti-recommendations, i.e. which books not to read. Often, I feel like there's a bias towards recommending books, after oneself committed to reading them. If I were asked to list the top 50% of books I've read (without counting), I'd probably list more than 50% of books I read.


This is fantastic! I've often come across book recommendations in HN comments and then struggled to remember where it was posted when it came time to purchase a book on topic X.

This also seems useful at aggregating different perspectives on the same book depending on the context of the thread it was mentioned in.


What an excellent UI. It's hard to make something relying on scraped data so nice and easy to use, this nails it!

Have a few books on the way now :)


Nice!

How is the site itself built? Is it a static page or do you have some sort of backend?

I did something very similar for a local podcast that gives book/movie/music/etc recommendations a while ago: https://enpassant.tk/ . It is statically-built with Ruby.


Hi lbrito!

The site is a Phoenix app (Elixir's popular web framework) with PostgreSQL. The pages are just Phoenix templates and I let Cowboy, the default HTTP server for Phoenix, serve the app directly... so it's not sitting behind NGINX or any similar web server that's frequently used as a reverse proxy.

I used Bulma as for the CSS, just to try something I haven't used before.

The only line of JS is in the dropdown menu's onChange tag, to submit the "form" when you select a book category.

I like the UI on En Passant! Really clean - great use of icons so I know what the media type is at a glance as I scroll.


I am trying to build something similar and I was wondering if there are any legal issues with showing content from the source website into mine, in your case for example the content is the user's comments.

In this page https://www.ycombinator.com/legal/ under the Intellectual Property Rights is written:

Except as expressly authorized by Y Combinator, you agree not to modify, copy, frame, scrape, rent, lease, loan, sell, distribute or create derivative works based on the Site or the Site Content, in whole or in part, except that the foregoing does not apply to your own User Content (as defined below) that you legally upload to the Site.

I really wonder if I need to ask for permission from each website about showing some text from their website into mine?


This is really cool! Is there any particular sorting on pages like https://mapfilterfold.com/articles/early_founders or is it just whatever order you put them there?


Hi, thanks! Frankly the articles (the three panels/links at the top of the home page) need more attention, but I wanted to have the feeling of "done" so I pushed the site out as-is.

On those pages, books aren't ranked in any particular order, other than trying to pull attention to some slightly buried books that I do personally recommend. I'd love to hear, are there any books you'd add to those pages?

The "Oddities and fun" page is simply a collection of books from my notes that stuck out as interesting while I manually approved all parsed comments and book mentions.


This looks great. I like how it is setup and I find the value it offers to be swell.

I will however complain that I shouldn't have to recaptcha myself if I took the time to enter my email address, find the email, and click confirm for it. What benefit is there to recaptcha-ing that?

My complaint is in hopes that they will remove that recaptcha for other users who happen to like their privacy and don't let their browser sell everything to mailchimp.

Edit: I will correct myself in that it looks like they may not have a choice according to https://mailchimp.com/help/about-recaptcha-for-signup-forms/


Hey myu701. Thanks for the link. I unticked the setting mentioned a bit lower on that page[0] to disable reCaptcha. It's not clear to me if that changes it for their landing page.

Testing an email alias of mine in incognito mode _looks_ like it's gone for me. Thanks again for bringing this up.

[0] https://mailchimp.com/help/about-recaptcha-for-signup-forms/...


Hey there. Thanks for trying it. I will try a personal email rather than my work email and see if I get the same result on the same machine.

Edit: It no longer recaptchas me, thanks!


This is exactly what a robot would say.

Clicking on the confirmation is only proves that you own that email. I could easily automate all of the steps that you said except captcha.


Ok...lets say I am a robot for hypothetical reasons.

So now a robot is on a mailing list. Why is that a bad thing? You may only want human eyeballs I suppose if you were somehow making a profit off knowing your audience was all humans a la broadcast radio advertising rates based on estimated audience size, but OP is using amazon affiliate links, not ads, so that argument is out of scope?


Would it be possible to create a similar list for the opposite purpose: Read these books to free yourself from the conventional Hacker News wisdom?


Looks good on mobile but on first load you need to unzoom for the content to fit.


Thanks! I removed the viewport tag. That should get it loading in an un-zoomed state.


This is very nice, thank you! A goodreads url in each book would be great


Thanks this is awesome!


Awesome. Thank you.


It says something about the disparity of intellectual depth in the community that Ayn Rand's Atlas Shrugged shows up next to books actually worth reading like SICP, GTD, or Cialdini's Influence.


I think you can strike lines between technical and fictional books. In doing so, I would expect a difference in quality since the avg HN lurker is a technical person.


Pretty much this. The average HN lurker will be reasonably sophisticated when it comes to technical subjects, but they are incredibly naive when it comes to all of those humanities subjects they skipped in university. When it comes to politics, economics, philosophy and other such subjects the average HN response vacillates between the extremes of either outright dismissal or arrogant pronouncements that are such a caricature of the Dunning-Kruger effect that you initially assume it to be satire.


> …incredibly naive [about] those humanities subjects they skipped in university.

Yip, but not nearly a naive as the humanity students are about technical subjects.

If we are going to fix anything, we need to force our arty friends to finish with some understanding of how the world works.

And by "how the world works" I mean the things that would remain unchanged even if every human vanished.

Like, to be topical, exponential growth.


This is great, too bad it's not getting enough attention. HN has changed in some respects.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: