Show HN: Meta book recommendations from Ask HN threads
mapfilterfold.com
mapfilterfold.com
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.
Hopefully that makes most of the bookmarking easier. I'll work on a solution for other pages later today!
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.
(4 months ago) https://news.ycombinator.com/item?id=21546850
(1 month ago) https://news.ycombinator.com/item?id=22419958
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.
Curious what's different under the hood.
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 also seems useful at aggregating different perspectives on the same book depending on the context of the thread it was mentioned in.
Have a few books on the way now :)
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.
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.
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?
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.
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/
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/...
Edit: It no longer recaptchas me, thanks!
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.
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?
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.