5 ms·

Ask HN: Is there a way to customize the stories on Hacker News?

27 points by anbardoi 2y ago

I just want to remove stories that link to sites like bloomberg, New York Times, Wired, etc. because they are paywalled/account-walled articles.

DevNinjaS 2y ago
There's probably not such a feature now. But if HN could provide such a function, I would wholeheartedly support it.
- chrisjj 2y ago
  Please let such a filter exclude videos and Twitter/X.
- empire_and_sun 2y ago
  Hint hint...your mental firewall will do all that very very soon. who's building that?
mtmail 2y ago
There's some tools to filter in the browser https://greasyfork.org/en/scripts?q=hacker+news+filter https://greasyfork.org/en/scripts?q=hacker+news+filter
chrisjj 2y ago
> paywalled/account-walled articles Most soon receive a comment having a bypassing link e.g. to archive.ph.
48864w6ui 2y ago
Hnrss.org and filter in your favorite way?
gtirloni 2y ago
This has recently been working for me: https://github.com/gtirloni/internet-serenity-now/blob/main/hackernews/tampermonkey_hn_posthider.js https://github.com/gtirloni/internet-serenity-now/blob/main/...
uraura 2y ago
I thought people here bypassed the paywall to read.
axiologist 2y ago
Wouldn't it be a more useful customization to have paywalled sites automatically redirected for being opened via https://archive.{is,ph} https://archive.{is,ph}, instead?
- anbardoi 2y ago
  Yes! This! This could definitely be done in a user script, or if someone's feeling especially bold, make a whole extension that redirects any paywalled news site to the archive.is equivalent
karmakaze 2y ago
I made hackerer.news[0] that filters all stories for the day into groups. I should add a paywalled group since I usually skip them too. Sometimes I'll just google part of the title and read a different article on the topic. It's rare for a story to only be reported by one source and I usually only care about deep details for tech which tends not to be paywalled much as business or politics. Edit: I just wanted to make an issue for this and realized this is still on GitLab[1]--should move it too. [0] https://hackerer.news/ https://hackerer.news/ [1] https://gitlab.com/karmakaze/hackerer-news https://gitlab.com/karmakaze/hackerer-news
gabrielsroka 2y ago
I wrote my own client ⁰ using the official HN API ¹ ⁰ https://github.com/gabrielsroka/gabrielsroka.github.io/blob/master/hn.html https://github.com/gabrielsroka/gabrielsroka.github.io/blob/... ¹ https://github.com/HackerNews/API https://github.com/HackerNews/API
khaledh 2y ago
I recently developed an extension to "dim" HN stories based on keywords and/or domain names[0]. You can also dim/undim stories manually. The list of keywords/domains is currently hard-coded (based on my preferences), but I intend to add an options page to set those. The reason I wanted to just dim the stories is to make sure I can still read stories in case something got dimmed as a false positive. [0] https://github.com/khaledh/hn-mod https://github.com/khaledh/hn-mod
- anbardoi 2y ago
  Nice work. I use Firefox but am gladly bumping this
kazinator 2y ago
You can use uBlock Origin for this! Go to My Filters, and add: news.ycombinator.com##a[href*="UNWANTED.DOMAIN.COM"]:upward(3) news.ycombinator.com##a[href*="UNWANTED.DOMAIN.COM"]:upward(3) + tr Maybe someone can come up with a better solution, avoiding the hacky upward selection and avoiding repetition of the domain match. The idea here is that we match an <a> element whose href contains the unwanted domain. Then upward(3) selects the great-grandparent element: we are assuming that the <a> is buried three levels within a <tr> element, and we nuke that <tr> element. Even if a nicer rule is found, it's still not great UX to have to go into the filter panel and copy-paste rules. Then we repeat the entire rule and add "+ tr" to nuke the successor also. The successor row has the small print: the story's points, who submitted it, number of comments, flag/hide buttons.
unraveller 2y ago
many user scripts for filtering sites on hn on greasyfork. search the url or site name there.
jpl56 2y ago
I'm still using an old extension called "Hacker News: Mark All Read" from 2015, it marks all articles on the page as read once I click on the button (at the bottom of the page), it's very convenient. It was updated un 2023 to keep on with new HTML formatting. Thanks to the authors (Guillaume Mouron, based on Daniele Mazzini code)! <3
andrew_zhong 2y ago
If you are looking for broader less heuristic based filter, you can try LightFeed (full disclosure, I am the creator of LightFeed) It filters and summarizes stories using LLM and your prompt on any site including HN!

27 points by anbardoi 2y ago

I just want to remove stories that link to sites like bloomberg, New York Times, Wired, etc. because they are paywalled/account-walled articles.

DevNinjaS 2y ago

There's probably not such a feature now. But if HN could provide such a function, I would wholeheartedly support it.

chrisjj 2y ago
Please let such a filter exclude videos and Twitter/X.
empire_and_sun 2y ago
Hint hint...your mental firewall will do all that very very soon. who's building that?

mtmail 2y ago

There's some tools to filter in the browser https://greasyfork.org/en/scripts?q=hacker+news+filter https://greasyfork.org/en/scripts?q=hacker+news+filter

chrisjj 2y ago

> paywalled/account-walled articles Most soon receive a comment having a bypassing link e.g. to archive.ph.

48864w6ui 2y ago

Hnrss.org and filter in your favorite way?

gtirloni 2y ago

This has recently been working for me: https://github.com/gtirloni/internet-serenity-now/blob/main/hackernews/tampermonkey_hn_posthider.js https://github.com/gtirloni/internet-serenity-now/blob/main/...

uraura 2y ago

I thought people here bypassed the paywall to read.

axiologist 2y ago

Wouldn't it be a more useful customization to have paywalled sites automatically redirected for being opened via https://archive.{is,ph} https://archive.{is,ph}, instead?

anbardoi 2y ago
Yes! This! This could definitely be done in a user script, or if someone's feeling especially bold, make a whole extension that redirects any paywalled news site to the archive.is equivalent

karmakaze 2y ago

I made hackerer.news[0] that filters all stories for the day into groups. I should add a paywalled group since I usually skip them too. Sometimes I'll just google part of the title and read a different article on the topic. It's rare for a story to only be reported by one source and I usually only care about deep details for tech which tends not to be paywalled much as business or politics. Edit: I just wanted to make an issue for this and realized this is still on GitLab[1]--should move it too. [0] https://hackerer.news/ https://hackerer.news/ [1] https://gitlab.com/karmakaze/hackerer-news https://gitlab.com/karmakaze/hackerer-news

gabrielsroka 2y ago

I wrote my own client ⁰ using the official HN API ¹ ⁰ https://github.com/gabrielsroka/gabrielsroka.github.io/blob/master/hn.html https://github.com/gabrielsroka/gabrielsroka.github.io/blob/... ¹ https://github.com/HackerNews/API https://github.com/HackerNews/API

khaledh 2y ago

I recently developed an extension to "dim" HN stories based on keywords and/or domain names[0]. You can also dim/undim stories manually. The list of keywords/domains is currently hard-coded (based on my preferences), but I intend to add an options page to set those. The reason I wanted to just dim the stories is to make sure I can still read stories in case something got dimmed as a false positive. [0] https://github.com/khaledh/hn-mod https://github.com/khaledh/hn-mod

anbardoi 2y ago
Nice work. I use Firefox but am gladly bumping this

kazinator 2y ago

You can use uBlock Origin for this! Go to My Filters, and add: news.ycombinator.com##a[href*="UNWANTED.DOMAIN.COM"]:upward(3) news.ycombinator.com##a[href*="UNWANTED.DOMAIN.COM"]:upward(3) + tr Maybe someone can come up with a better solution, avoiding the hacky upward selection and avoiding repetition of the domain match. The idea here is that we match an <a> element whose href contains the unwanted domain. Then upward(3) selects the great-grandparent element: we are assuming that the <a> is buried three levels within a <tr> element, and we nuke that <tr> element. Even if a nicer rule is found, it's still not great UX to have to go into the filter panel and copy-paste rules. Then we repeat the entire rule and add "+ tr" to nuke the successor also. The successor row has the small print: the story's points, who submitted it, number of comments, flag/hide buttons.

unraveller 2y ago

many user scripts for filtering sites on hn on greasyfork. search the url or site name there.

jpl56 2y ago

I'm still using an old extension called "Hacker News: Mark All Read" from 2015, it marks all articles on the page as read once I click on the button (at the bottom of the page), it's very convenient. It was updated un 2023 to keep on with new HTML formatting. Thanks to the authors (Guillaume Mouron, based on Daniele Mazzini code)! <3

andrew_zhong 2y ago

If you are looking for broader less heuristic based filter, you can try LightFeed (full disclosure, I am the creator of LightFeed) It filters and summarizes stories using LLM and your prompt on any site including HN!