uBlacklist: Blocks specific sites from appearing in Google search results | Hacker News Reader

uBlacklist: Blocks specific sites from appearing in Google search results

github.com

239 pointsst_goliath89 comments

I have said this before, will say it again. This can be fully achieved with uBlock and custom filters like the one maintained by quenhus[1] that can block websites across multiple search engines. You don't need to increase the vulnerability surface by installing yet another extension on your browser.

[1] https://github.com/quenhus/uBlock-Origin-dev-filter

But you can easily block websites from the results page.

Is it domains or just pages?

Domains, with a wildcard option for the subdomains.

How do you use ublock to block domains on the fly as you search?

This seems to be targeted against copycat spam of known sites. I just want to remove some sites from my results because I hate them.

You can generate a filter for that here:

https://letsblock.it/filters/search-results

The repo you’re linking to seems to itself state that uBlacklist is “more efficient than uBO in this case”

Almost any blocking extension can be achieved with uBlock, that doesn't make them any less useful:

- a dedicated extension offers a better experience when blocking (uBlacklist for it's case certainly does) - having everything in uBlock will disable all blocking, even if you just need some disabled (in this case, you might just want to see all results, but keep blocking tracking + ads)

It might still be a good trade-off to just use uBlock for the vuln. surface, but there are good reasons to install another extension.

You could just uncheck the list from quenhus when you want to get all search results. I don't think thats much harder than disabling an extension.

I'm referring to built-in disable buttons from the appropriate extensions

That page itself says:

uBlacklist (more efficient than uBO in this case)

Google used to have own extension for chrome called "personal Blocklist", which would allow users to block specific domains from eve appearing in search results. It was great to remove instagram/LinkedIn/quora and various stack overflow clones.

But as all good things, it was killed by google

https://killedby.tech/google/personal-blocklist/

Google used to have this build into google, directly. It’s near immediate discontinuation made me realize that google was on a downward slope of profit vs quality results.

This made me realize how much I hate Google. It mutated from something good and useful to something negative. But it's representative for pretty much everything in our time. Google search on the desktop is just awful. The UI looks like it's made for mobile. Instead of search results you receive specially formatted "AI results" that are just in the way all the damn time, add nothing worthwhile.

Old search I liked actually had results where you could find out what the link target was.

And all results are Reddit or Wikipedia or 10 best things. As there's no other content on the internet. They also don't monetize any other content, because for them forums aren't content, audio isn't content, apps aren't content. They're driving this dystopia forward. And they want the made nest. In development aka new sites aren't welcome.

Google has turned into a pos company with too much power and is spying on everyone.

Their attitude towards their users and customers is hostile.

This ultra capitalism needs to stop. In fact I believe Google needs to be disowned. Not just Google. All bigger than the state tech companies, Microsoft, Meta, Apple. The profits used to better the living standards of everyone on the planet. Where is the money going, the profits? Not into the 99%'s hands.

Those hostile to the population companies need to be disowned.

Agree. I am so sick of everything turning to shit. profits....bullshit....each one of these so called profit chasing decisions were short sighted and cost them billions and billions and possibly trillions of dollars had they just continued to care about the product and what it was originally meant for

We should compile a list of Google search features that were killed.

The other one I remember is live results from Twitter. And then "cached" link prominently showing (now it is well hiddenif available).

What other usability features has GSearch removed ?

Just switch to Kagi. Google is, alas, over.

As of this morning I have fully deGoogled my laptop. Kagi, Firefox, Infomaniak kSuite. Based on my research, with more than 500GB of storage needs kSuite is the only Google Suite replacement with a competitive offer. Proton seems to only offer additional storage space in the form of 500GB per business account for 11 EUR a month. This is like 4-5 times the Google price. mailbox.org Premium is €0.20 per 5GB or €40 per TB per month. For Standard it's double. kSuite is 6.58 EUR a month for 3TB so I went with them. Signup was not smooth but once that was done it is actually working quite well. One snag: be careful with your domain transfer, back up your existing zone file before transferring, I've not considered this but your old provider will delete the records. Luckily Namecheap support found a backup for my domain despite chat claiming it is gone.

Hi! Note that you can have 500GB storage on an individual Proton Unlimited account too: https://proton.me/mail/pricing, while our Family account comes with 3TB total storage: https://proton.me/family. In addition to that, all paying users get storage bonuses for each year of their subscription: https://proton.me/support/storage-bonus.

Hi! Let me emphasize:

> with more than 500GB of storage

https://proton.me/support/increase-storage-space suggests for this case

> For example, if you wanted 2 TB of storage, you could sign up for Proton Business with four users (4 x 500 GB = 2 TB).

Thank you for spotting this, we will update the article as soon as possible.

Family account is 29.99 € /month for 3TB with no clear way to add more storage.

Meanwhile, kSuite is 6,58€ / month for every 3TB. And email storage is unlimited.

I am not an Infomaniak employee, I am merely deGoogling here.

A comment suggesting kagi and detailing how they are happy with it in 3 2 ..

That's what I was expecting too. I'm very happy with kagi...

You got me. But seriously...

I never managed to get it to stick, $10/mo for a search engine seems like a lot, when compared to the free DDG, and from what I remember the results weren't impressively better, just maybe a bit better.

Yeah, $10/month is too steep and the only reason I’m not signing up. I’d sign up for $3.

> $10/month is too steep and the only reason I’m not signing up. I’d sign up for $3

I suspect Kagi aims to corner the premium market. The 300 searches will probably get discounted. But I don’t think they should nor will discount unlimited searches + AI.

I'm not going to play the guess your usage game.

They can give me the best rate with a cap or I'll wait for the competitor to show up. Duckduckgo is fine enough for now.

I feel the same way, I'm not going to count my searches, and I don't want to start guessing whether this search is important enough to use Kagi for.

Nobody wants a three dollar customer. They'll complain, ask for special features, do chargebacks on their payments etc.

I tried kagi for a bit, but these days I think phind gives the best results for cs related searches. In addition to good search results, it shows llm responses with a chat interface to continue that chain.

https://www.phind.com/

Phind gave me really underwhelming results when looking for more technical topics, where it misses the point totally.

Haha. But seriously, Kagi gives you all this out of the box, in any browser…

Is it still doing well? Getting better? Worth paying for?

I think it's doing well. I've been paying for it since June and I haven't looked back. For me, it's been either equivalent or better than Google for work-related (programming) searches.

Getting better, definitely. I've been using it since beta, and while in the early days I'd use !g often to switch to Google, it's been months since doing that has made a difference. These days, if Kagi can't find it, Google can't either.

A happy Kagi customer reporting in.

Guilty as charged… it’s just such a great experience and the incentives align.

While I appreciate the author's work, I suggest ditching Google Search altogether in favor of DuckDuckGo or anything else. At this point I only use Google when I need to buy something, as Google has become an expert in finding products tailored to my needs (but not in checking knowledge/facts - my go to here is ChatGPT/Wikipedia)

This works for DuckDuckGo along with several other search engines listed on their Readme. I suggest using SearXNG instead as DuckDuckGo is also a problem now: https://www.wired.com/story/duckduckgo-microsoft-twitter-ft-...

It works with DuckDuckGo as well. I like the privacy focus of DuckDuckGo but it doesn't seem much better than Google when it comes to promoting garbage sites like slant.co over information written by a well informed human being.

Exactly. It actually feels to me that it does worse than google in this regard.

How is this relevant to the post?

It's just that Google search results are now full of ads, so I suggest changing the search engine altogether and not just blocking certain sites. Does it make it more relevant to you?

ChatGPT is the last place you should go for facts. It's a language model, not a data model, it only knows what words usually statistically follow another one, it doesn't know anything about facts. Granted, it's very good at sounding like it does, but you never know when it hallucinates.

I am aware of this. But my daily questions these days revolve around TypeScript or something as silly as "what is the average weight of a cat". Back in a days it took like 10 sec to get this from google, but now it's just a pain in the back to open all the websites, accepting cookings... and you know the drill.

Use FastGPT by kagi search. All its statements come with the web source snnoted to research or verfiy the claims.

https://kagi.com/fastgpt

While better, it still gets things wrong.

Seems more accurate than most Google search results

Same with google search.

I'd like to use the opposite approach: top-list search hit results on websites that I've bookmarked another subpath of or visited before. Difficulty being: when to stop requesting later result pages to get a hit. Both Firefox moz_places and Chromes Default/History sqlite databases have the info to populate such a list.

Except that can be only implemented on the search-engine side, obviously.

If you're currently into researching sqlite questions and it happens you prefer one sites explanations, it likely comes up in the first few pages of a result of new questions on sqlite. By looking into bookmarks and history the users preference is clear.

In general you're right as in: most often in the first few result pages, there is no hit of sites you prefer. I think an extension that lists the most popular pages that have user-agency to the top can already help without curated blocklists: rank wikipedia, stackoverflow sites, github and reddit to the top.

There was a Show HN for "hypersearch" in 2022 that did something similar but in a sidebar - https://github.com/abhinavsharma/hypersearch

Kagi provides a feature to boost or even pin certain domains.

I'm looking for something like this, but for YouTube search results / recommendations

You can do it easily using ublock origin. Just select the div of the video you dont want to see anymore, and append to the filter string something like this

:has-text(/channel name that I hate/)

I filter any video with less than 1000 views, since youtube started to push insanely small channels to my recommended videos lately. This is the filter to accomplish that, on mobile youtube:

I have so many filters on my ublock youtube section that nowadays their algo doesn't seem to even bother me so much. I hardly let Youtube show me anything that I would not really like to see.

Another anoying thing that started showing up on my feed lately were very old videos, 5, 10, 15 yo videos, with millions of views. Clearly meme worthy garbage. So I started filtering out any video with more than a million views from my feed using the following filter:

m.youtube.com##ytm-rich-item-renderer.fresh-feeds-dismissals:has-text(/M views/)

I never, ever, watched a single "short" youtube video in my life, so maybe their algo tried to memefy my feed using this alternative path. No way, my time is precious and every single video I watch must add something.

To remove "Sponsored", i.e. videos that are really ads, from your feed:

m.youtube.com##ytm-rich-item-renderer.fresh-feeds-dismissals:nth-of-type(1):has-text(/Sponsored/)

My youtube feed nowadays only shows videos that have between 1001 and 999999 views. To me, its the sweet spot of quality for youtube videos. If you didnt catch the upload on the first week, you'll get it down the line, when it passes the threshold of 1k views. I know I maybe not getting a Veritasium video suggested to me once in a while, since nowadays he tends to get millions of views on the first day already, but once I remember about those channels I really like, I'll manually visit their page and retro watch all the videos I missed in between.

Ublock Origin is the single most useful piece of software one can use to keep their cognitive sanity in the modern world.

That's some great tips thanks for sharing.

I've gone nuclear and just have my home perma blocked using newsfeederadicator extension (works across a few sites), and shorts blocked in subscriptions via a ublock origin rule.

Are you actually able to use home / reccomendations using this solution (I opted out of both a long time ago for many of the reasons you state), or is it just to fix the terrible search results yt gives?

The full css path of the div you want to block will be different on every type of list, "home", "recommended", "subscriptions" etc. But the basic string rule will be the same, you just have to set 3 or 4 sets of the same rules and you're done.

Also, you'll have to set the rules on both mobile and desktop, since the layouts are different. But after setting up the first rule set, it's just a copy and paste job for every other list. Not a hard task at all. I haven't touched my filters in over 3 months, the last time was to add the "no video with more than 1 million views" rule.

These rules basically never break, I only have to edit the filters when the Youtube algo decides to push another set of annoying videos on to my feed. The layout, it seems, never changes.

I use Channel Blocker extension[1], it can block channels and individual videos.

[1] https://chromewebstore.google.com/detail/channel-blocker/nfk...

It also works on other search engines, which is good to know.

And it has blocklist subscriptions, but none selectable by default.

It would be nice if someone could make a blocklist to make Google or DDG similar in low spam to Kagi.

Pretty good stuff.

Some people already maintain some lists that you might find helpful [1]

[1] https://github.com/rjaus/awesome-ublacklist

Blocking medium.com has been the best thing I've done with an extension in 2023. Thanks uBlacklist!

It’s pinterest for me. It’s infected so much image search space and never takes you to the actual source images. It’s infuriating.

I'm using it since... uh, forgot how much time, anyway it works great to filter out the Pinterest junk that invaded Google Images a few years ago.

I just wish there was some similar extension to filter out Ebay sellers from search results based on their shop/user name and/or feedback ratings. Ebay blacklist likely works only on user names, and I'm not sure it is effective when searching as a buyer myself.

+1 to excising the internet cancer that is Pinterest

Curiously, I was just playing with it yesterday.

In the end I decided to just replace it with uBlock origin plus some filters [0], [1], [2].

While this solution doesn't add a convenient button in the UI to show the hidden domains and to quickly block them, for some reason this extension didn't work if I had javascript disabled, while the native uBlock one did.

I believe there's also a userscript that does something similar, but I never tried it.

[0]: https://letsblock.it/filters [1]: https://github.com/quenhus/uBlock-Origin-dev-filter [2]: https://github.com/rotgruengelb/BlockModReposting

Google used to have this feature built-in, I dont understand why they got rid of it.

I am sure there are many organizational reasons, but the mere existence of a first-party “fix search engine results” app implies that personalized search must be broken for enough people to warrant its creation.

The sites I am most likely to exclude are the most ad infested. Seems a direct conflict with the bottom line to only show high quality content.

We all "know" deep down why they removed it.

Deep down all these apps and websites are glorified databases of "something", and giving the users more options means:

1. Users find exactly what they want, more quickly.

2. They spend less time "browsing" or windows shopping to use a physical world term.

3. The app has more control over what they show you, so they can sell that feature to others.

4. Users have less perfect information, so they never know the perfect or optimal choice or purchase. This again means they can convince producers to join.

5. Less direct access to their database so less scraping.

6. People don't realize how simple the data is, just a bunch of Metadata really.

All of this is given an aura of legitimacy and user friendliness by complicit or unknowing growth hackers, marketing people, UX designers, engagement people, and dozens of other names.

Interesting. Looks useful for the many sites that refuse to work unless I turn off my stalking preventer (or ad blocker as they call it) or have a metric shed load of clicks to not opt in or accept "legitimate interest" (will save me hitting back / close every time I see the "Admiral: Privacy and Consent Management Platform" logo, amongst others).

For anyone looking for something a tad more user-friendly, I use the extension Google Hit Hider to accomplish this. Works great. You can sort and filter the list of blocked URLs, de-duplicate, and do all sorts of useful manipulations. Works well on DDG and Google, haven't tested other engines.

This is way cooler than what I am using, in fact a couple of weeks ago I asked X/twitter about a solution to this same issue and someone suggested this: "Highlight or Hide" on the Chrome extensions store - which is actually really simple but more manual work...

Bye bye Quora!

Was thinking the same thing

Whoever at Google who is boosting quora is a real bastard.

Quora, pinterest, medium: Surely they have good million dollar contracts with Google to give them a punch in search results.

Must be a new form of advertising - paying people to add comments to news articles with the word "kagi" everywhere and some sort of anti-google phrasing.

You used to be able to do this right on Google’s search-results page.

Taking it away was just another “up yours” from Google to you.

There is also letsblock.it I found it through HN and it has worked well for me.

this should be defacto in google, quora.com should be blocked 100% from google search results, so does all the payment requiring news websites like ny.com, but they aren't because they are all the same..

Or just switch to Kagi, it's far better than Google anyway.

Mmh.. Negation curation.. Can this work against seo?

This is asking for permission to access my personal data on more than 190 sites. I don't think so, buddy.

It's specifically asking to access all the Google sites with a different tld, a far better approach than asking to access every website with a single permission

The lack of this feature on Google shows whose interests they truly have at heart.

the REMOVAL of this feature from Google…

There might be other explanations, thought not necessary better. For example maybe they are following Steve Jobs' dream of single button apps? All not essential features confuse people[0]. Google have already hidden "cached" version link in tiny hamburger menu.

Or they assume they know better what the users wants than the users themselves. YouTube straight up ignores dislikes - and will show disliked video in mixes or at first place of search results.

If we give Google the benefit of doubt, maybe people just kept accidently blocking useful sites and later complained they couldn't find anything? Thought Steam deals with that perfectly: search results start with disclaimer: "8,875 results match your search. 2,053 titles have been excluded based on your preferences." - and it takes 2 clicks to go back to unfiltered results.

[0] - well, at least Gnome project follows that idea, killing any discoverability

Reply on news.ycombinator.com