Pinterest does a whole lot specifically bad - they tend to trap users with login prompts on entry and make it impossible to see the full image. If I'm searching for a meme to DM to someone or a photograph of a landmark Pintrest is something I avoid like the plague. It's essentially a black-hole for images - people pin them and then they are lost forever.
I like Pinterest, it is a useful site. But it is basically a (image) search engine itself, so it's pages shouldn't be in Google results. Just like you don't want Bing or DDG image search results showing up in Google searches.
Pinterest could simply tell robots to 'noindex,follow' the page to help search engines understand there is no utility in it appearing for a search. But agree, perhaps it shouldn't rank first anyway.
Guess it depends on whether it's UGC entirely or whether the CMS could note these kind of pages. Again, shouldn't be hard if it's just a simple link as the main content of the page.
There's precedent that would make it in Pinterest's best interest. IIRC Google's "Panda" algo update basically wiped out search referrals to certain content farms overnight, can't recall specific examples but there were some at the time. They were labelled as low quality and subsequently faded into obscurity.
//added
EHow is an example of a site that was flagged
Yeah MDN is the de-facto reference guide for web frameworks. W3schools feels like beginner tutorials. For me, I want “what happens if I pass undefined as the second arg to this function” and w3schools doesn’t answer that. It’s fine as a site, I suppose, it was insane when it outranked more complete websites. Off topic I see this more and more with Stackoverflow where semi-related questions outrank official docs.
It often outranks MDN official docs which these days are very readable, in-depth, with examples, and a simple UI and the only time I click to inappropriately named W3Schools is by accident because it's the top result and I'm expecting MDN and then I click back and that's why I don't like it.
I have similar feelings about baeldung.com. It isn't that the content is inherently bad, but it is super annoying that when I search for a java class the first few results are from baeldung instead of the official javadocs.
How wonderful to differ - I actually love bite sized, "tldr" or baeldung and almost always get what I need. (not to mention that quite often javadocs of many project are just empty or severely lacking... )
Yeah. Search for class, open official docs, read very short description. And still have absolutely zero idea is where this class is used, and how it should be created, and what alternatives.
Nothing about that makes w3 a poor source. I get why you block it - so MDN is shown first. That’s fair. Im just saying its not a fault of W3. Because I tend to agree with that guy. W3 really isnt bad. A bit simplistic, and MDN is better, but its not bad.
Yeah the simplicity is what I prefer about W3schools. A lot of times I want a quick code example and don’t really want to go in depth about how something works.
Yes, and historically those code demos were often wrong in many important ways (like not working in all browsers, or insecure). They are much better now but reputation is hard to recover.
That must have been like 12 years ago or more at least because in my career I don't have much memory of feeling betrayed by the information on their site.
It's a question of presentation. Back then (and to be clear, I'm talking about 00s here) the correct way was to point out that something might not even be doable in cross-browser way at all, and listing the proprietary features that'd let you do it in specific browsers while clearly labeling them as proprietary. Quality websites would do that, while w3schools and the likes would just give you the snippet for (usually) IE. Worse yet, in many cases they would keep suggesting the old proprietary approach even after a new standardized way of implementing something showed up.
That's why Kagi's personalization is so cool here. Kagi doesn't have to block it, and neither does the user. The user can just lower it, and raise MDN, because that's what they want in their search, and it helps them more.
From what I understand a lot of it is residual. W3schools has always been directed at beginners, but it also used to be wrong rather frequently.
This site [0] used to be the hub for the hate and now has a message saying that w3schools has gotten better. If you're curious, they still have a link to the web archive that has all the mistakes documented.
The most egregious of their mistakes to me:
> […] professional web developers often prefer HTML editors like FrontPage or Dreamweaver, instead of writing plain text.
W3Schools was essential in helping me get to where I am today. Learning HTML and PHP through it was great, my favorite feature being their "Try It Yourself" button, which opens up a sandbox where you can play around with PHP, HTML, CSS, JS etc (note it's not just client-side languages like MDN does). For beginners, this kind of thing is essential, and this was before MDN was good. And even now, I compare MDN vs. a W3Schools page, and the overall design and content is more welcoming for beginners.
w3schools used to be one of the best resources ~15 years ago. Today I intuitively avoid it because
1. The ads make the fan on my laptop spin (I prefer to avoid ad-heavy websites over using adblock)
2. The 'Try it yourself' button is fake. It doesn't actually run the code and doesn't let you edit it.
3. It is not a good reference. To give an example, the last time I visited w3schools was upon searching "react ul li", which landed me on the page https://www.w3schools.com/react/react_lists.asp . Ironically, the reason why I was searching the term, was to find advice on setting the "key" prop in on list items. The w3schools example doesn't even mention the prop and produces errors. Compare that to the official reference which ranks much lower in Google search: https://react.dev/learn/rendering-lists#where-to-get-your-ke... .
As others have said, it's t least in part because of the beginner orientation. I'm not a web developer and I found it useful when looking up some html/css stuff for a project years ago. But for example in python I find it infuriating to see geeksforgeeks or whatever at the top of my search results. I want SO or documentation for most stuff and consider the rest spam.
All that to say, I think beginner oriented well SEO'd content is going to be polarizing, which makes Kagi's approach great for everyone, if you want it it's there, if it annoys you, quickly block it
Yep. W3schools is one of those sites that helped me begin my career. I still find it useful to look up something in a hurry. No idea why so many people hate it.
I learned from w3schools in the early 2000s, and helped instruct my high school web design course with it.
I wish I hadn't. There were many innacuracies, outdated advice, and over simplifications.
I came up on a slew of weird and inconsistent sources, and that bit me when discussing things in college. I often felt over confident because of these type of sites.
W3schools (used to at least) gave a very shallow intro to programming languages (PHP, ASP, etc), and I remember a professor basically laughing at me once because of that misinformation.
So, maybe others had experiences like mine.
EDIT
They also had these weird "certifications" for technologies that were basically really dumbed down quizzes. They were made to make people feel special, I think. Dunno, I did them all, and I almost got laughed out of class.
But like what? Genuinely curious for an example of "blatantly false claims"? As I'm on the "right" side of the spectrum, and these kinds of comments come from a criticism of "misinformation".
Also, as an aside. What is a "false claim"? If it's a claim, they're asking you to believe it, which means they need to motivate it and provide details to back it up. Seems precisely where all the media generally is (including CNN/BBC/etc), pandering to peoples' bias without saying anything factually incorrect.
This feature is Kagi's single best innovation in search. It's what makes their results so much better than Google's: where Google's results are usually drowning in garbage websites, Kagi comes built-in with a great set of spam filters, and on top of that you can adapt your own personal list of good and bad domains as you go!
Didn't Google have this feature like two decades ago (blocks, at least)? It's not quite a search innovation, more like a "caring about your users as a business strategy" innovation.
If you think through the feature, it's not at all obvious that it's a good idea for an established search engine targeting the general public [0]. But even then it might be a great idea for a search engine that's trying to establish an initial niche. It doesn't even matter whether the idea actually works well in practice: if 100k people think that manual curation of search result domains is a high value feature and are willing to pay for it, the feature has done its job.
My ML prof in university used this as an example for something that could happen but is highly unlikely. Each particle of oxygen can be in either half of the room so it's like 1/2 to the power of the number of particles (avogardo?) chance that all the particles are in one half suffocating the ppl in the other half. He was such a great guy.
This is a good point. Did Kagi ever consider having a freemium tier that uses adverts to pay for the (search) content? I don't think it is a terrible idea if it helps Kagi to improve their paid search.
As long as they don't take outside investment that's not a certainly. People complain about the prices but it's charging the real cost that let's them be able to offer something good without mortgaging the future.
I concede that if they got big enough, someone would come with enough money that they couldn't refuse, and buy them to try and "unlock vale" by racing to the bottom.
Capitalism is the only reason they exist in the first place.
Don't confuse an entire economic system with the unique problems caused by public ownership or investing (i.e. growth at all costs). Plenty of smaller organizations would be quite happy to operate with modest profits for their owners and never scale beyond that point.
> Capitalism is the only reason they exist in the first place.
Precisely, which is why enshittification is a when, not an if. A paid search engine will never pick up enough users to survive at $10/mo. It will either be acquired or start throwing in "non-intrusive" ads to balance the books.
> start throwing in "non-intrusive" ads to balance the books
Here is a personal guarantee that this will never happen with Kagi while I am in charge.
There are too many ad-supported search engines and browsers out there. I would not waste 10 years of my life to make yet another one. If Kagi can not sustain iteself with memberships, it will be the end of it.
The initial reason I started Kagi was to provide my kids with an opportunity to grow up in an ad-free experience of the web. In this interview [1] I talk more about my motivation.
How has this broken any privacy policy? To me this seems like general information that can’t possibly be traced back to any individual users or groups of users.
"Enshittification": God, I love this term. It so perfectly captures what happens to a product when revenues stall and product managers think of anything to drive new revs. See: Microsoft Windows or Microsoft Office.
A key point is that Kagi raised its prices already so it was sustainable. The reason most services get worse is because they are downright unprofitable and it is impossible to get them there without either making them drastically worse, or raising prices, which free sites can't do.
Kagi has already put itself on a path where it doesn't have to make itself worse to cover its expenses, because it is already covering them.
No, the main reason why companies turn to shit is because they’re publicly traded and accept VC money so sooner or later, they’re beholden to shareholders demanding infinite growth and profits.
This is honestly probably the most tiresome argument that constantly pops up in every HN thread.
Imagine if one of your friends come back from a holiday trip and tells you how fantastic the place they visited was. Is your response "Well, maybe that place ain't going to be so nice in a few years!"?
The usefulness of a per-user block list is not fighting outright spam. The statistical filters, etc, as applied by the search engine, work best here, as you correctly state.
The point is to remove certain well-known sites with information about well-known topics, which is unhelpful for the particular user, me. That information is technically not spam, but it's on a wrong level, badly presented, etc.
That is, if I search some info about a particular CSS trick, I want MDN be on top, some professional blogs be near top, and w3schools be on the third page. For someone else, priorities may be different, but I know what to expect for the search queries I issue, and I'd be glad if the search engine helped me.
Equally, certain domains are known to be more rich with information I happen to seek, like reddit.com, so I'd like to give them a boost. Some other people may prefer something else, or not give any boost at all.
So, this is personalization, not anti-spam. If Google kept it, it could even bring important insights into what ads I would skip and what would click :)
Maybe 90% of users would ignore this feature. But the 10% of power users maybe disproportionately important: with more disposable income, more clout on social networks, more decision-making power. I suppose that the shutdown of Google Reader, which affected a relatively small number of users, affected all the wrong people, and destroyed so much goodwill that the losses from it are many orders of magnitude larger than whatever savings from not supporting it any more.
> That information is technically not spam, but it's on a wrong level, badly presented, etc.
I'd count that kind of low quality site as spam, and the problem is that there's a basically an infinite supply of them. The boost feature is more interesting in that the curation is at least not a never-ending task. New high quality sites that I know about don't pop up all that often!
It's still not an entirely robust feature since you'll need to do a lot of quality work to prevent mediocre results from a boosted site from shadowing better results from unboosted sites, especially for users with a ton of boosts. But that quality work is much easier to justify when the manual curation is one of your core selling points and used by a large proportion of the customers vs. when it's a feature in a dusty context menu that basically none of the userbase interacts with.
> Maybe 90% of users would ignore this feature. But the 10% of power users maybe disproportionately important
I'd be shocked if the number of users for this kind of feature was anywhere near 10% of all search engine users. Something like 0.1% seems much more likely. But again, if those 0.1% really want that feature it's a great fit for a power user search engine.
It still doesn't mean that it's a feature every search engine should obviously have.
> you'll need to do a lot of quality work to prevent mediocre results from a boosted site from shadowing better results from unboosted sites
That strikes me as a non-goal. As the search engine, you can't know what attributes a user is valuing when they boost, and so you can't know when it's appropriate to override that boost. "Better" is subjective, contextual, and comes in thousands of different flavors.
> New high quality sites that I know about don't pop up all that often!
But when they do, how are you going to know if all you see is the mediocre spam instead? No, I want to block the domains which are never going to be a good result. This may actually surface things I will want to boost.
Pintrest is among the top blocked. It clearly doesn't have spam, but it commonly shows up because of it’s vast array of user content. For people not interested in Pinterest content, this is a very useful feature.
NYTimes is another good example, great content for subscribers, essentially useless spam in a search result for non-subscribers. The only person who can configure this is the user.
They used to have outright incorrect information a long time ago. They fixed the site and are much better, but it's aimed at a non technical audience in the end.
Is w3schools still that bad? I know that it used to be awful in the past, but IIRC they managed to do a complete 180 in this regard. I'm not into webdev but I occasionally check how to use a particular feature on there and it always delivers a pretty good reference.
w3schools garnered a reputation for stating outright wrong facts. I think they rectified that part, but I think the way they break down information is still far inferior to MDN on basically every topic.
This is a well written post. I describes me perfectly and my knowledge of HTML/CSS/JS. I usually start with w3schools, then if I need deeper info, I repeat the search with "mdn". That does the trick!
And this is exactly why the custom domain block list is so useful. Blocking w3schools entirely would be unhelpful to beginners, but a lot of experts find their content to be less helpful than that on other websites and would rather not see them at all.
Totally get where you're coming from about removing certain domains from Google search. But let's take a step back for a sec and consider what this could look like in a different light. Imagine if Reddit got the axe - a site that, despite its redesign snafus and sometimes annoying mobile login walls, a lot of us here on HN still find pretty useful.
Now, I'm not blind to the fact that Reddit's user demographic has a lot in common with us HN folks, while Pinterest... well, not so much. It leans towards a non-techy crowd, with more women users. But the way I see it, if Google keeps popping Pinterest up in search results, someone out there must be finding it helpful.
Perhaps what this all boils down to is a bit of a demographic reality check, rather than a definitive statement on any one platform. And just to be clear, I'm not standing on a soapbox championing for Pinterest, but I reckon we could probably nudge Google to get a bit smarter with its algorithms.
So, what if instead of giving domains the chop, we pushed for Google to up its game on the personalization front? Like, if I'm consistently booting Pinterest from my searches with "-pinterest", Google could take the hint and drop it down the ranks for me. Wouldn't that make more sense than completely blacklisting sites that some folks might still want to use? In the end, we all win if Google gets better at tuning into our search habits, right?
I'm not saying we need a global chop/removal, I'd be happy if i had personalized filters for me (using my logged-in account + removal settings) to remove pinterest just for me. Same for Quora and ExpertSexChange or whatever that other technical site was.
Same for youtube, a simple blacklist would make my search results much nicer, but youtube seems to push a few large media companies as hard as possible.
> Third, some people will probably block domains they shouldn't have blocked, and then have a bad user experience in future searches as the sites with genuinely best results is blocked. And then you're only left with only bad options: ignoring the users' stated preferences which they'll hate, or serving bad results that they'll also hate.
That's a real possibility, but it can be addressed by presenting a block that would say, for example "results from blocked domains" or some such.
People would use it to block all the spam sites that host copious amounts of (potentially Google's) Ads/Analytics, and since they might actually start finding what they're looking for this will reduce engagement on the search results page and with the sponsored links.
This is bad for Google's bottom-line thus why such a feature was removed and will never be coming back.
when google first came out it had nothing to do with that, it was that they offered as strong algorithmic ranking in the feed. Before that the search providers ran in a few ways (by my memory):
1. Sell the first few results but make it look organic (this all happened behind closed doors and I knew a few people who paid for such results)
2. Hierarchical/topic based collections of links (This is what Yahoo was famous for in the 90s if I remember correctly, I'll keep saying that because it was a long time ago!)
The whole "google cares about you" was an outcropping of the organic flocking to google because of their reliable and algorithmic results that just seemed trustworthy, they capitalized on that and it honestly lasted a lot longer than I expected before they really started being evil.
Yeah. Engineers hate when something isn’t a technical feat. The best innovations are usually software people working with the right people from other disciplines to make a thought out CRUD app.
Firefox on Android only supports a handful of curated add-ons. You can apparently create your own collections or install anything with Firefox Nightly but I've not done this.
You can do that with Firefox Beta too, which likely will give you better experience / stability. It's a shame you have to jump through so many hoops to get it working.
Yeah, this is _the_ reason I started using Kagi. The default results aren’t significantly better than Google/Bing/DDG, but being able to block and boost domains is a killer feature.
I was just about out of my free searches on Kagi before I ever realized it could do this. That feature converted me to a paying user (although I've got browser plugins to block domains from Google search, but boosting is something I can't do there).
Yeah, boost and pin are just as important to me. Google inexplicably doesn't show Wikipedia on the first page much of the time anymore, and I can fix that with Pins in Kagi.
I have been using this with ublock for a long time. It actually makes me wonder if that is the reason I dont have the complains about google search that I read about here
This is the feature that I originally started using Kagi for, but I wouldn't say it's the best innovation. Their default search results are already dramatically better than Google's, and their AI tools which can summarize individual results or give you a quick answer for your question based on the first page of results are way more innovative. All of those features combined and their subscription-based business model make Kagi by far the most innovative player in search at the moment.
Also: incentives are aligned. Kagi gets money from its users, so it will do what's best for them. Google's incentives are more muddled, to put it mildly.
Simple, common sense features for any search engine that actually wants to help you find what you’re looking for. Such a stark difference from what Google offers.
The top-ranked sources really show Kagi's current demographic. It obviously is attractive to programmers and IT professionals.
I wonder if they are interested in attracting users with more diverse searching requirements, and if so, how they would be able to actually reach them.
> It obviously is attractive to programmers and IT professionals.
I would say this is a blessing as it allows us to focus on things that matter to tech-savvy users (Kagi founder here).
> I wonder if they are interested in attracting users with more diverse searching requirements,
Yes, we are :)
> if so, how they would be able to actually reach them
Through our family plan [1], most likely. This is how I got my family members on board a paid search product.
I might add that "Kagi for Kids" is the first true attempt at having a search engine made specifically for kids, with a host of parent controls, which was never a thing in the last 25 years for some reason.
We are also seeing more and more users who are not techies (but are tech-savvy) becoming members. This may be followed by non-tech savvy users in the future. I do see the first signs of the age of ad-supported, "free" search, coming to an end [2].
I was hesitant to configure my partner’s devices (family plan) to set as default engine but eventually convinced them and they haven’t even brought it up once since. I assume no news is good news! I would be less hesitant to recommend to other non-tech friends/family if it were easier to configure as default on Safari. Heck, I’d even entertain gifting subscriptions. The quality of results for me has been incredibly reliable. Enjoyable, even. And, I didn’t even take the time to start configuring these kinds of rules. Love this. Great product. Wishing long, lasting success.
Unfortunately we can not influence the choices that Safari makes, but we did pull an effort to build an entire browser [1] to replace it because of that :)
We haven't begun work on the Linux version of the browser yet. We will need to significantly expand our team before undertaking the development of a browser for a new platform.
The macOS browser is based on WebKit. If the team can’t justify the Linux version yet, changing it to a whole different engine (and one which is notoriously hard to embed, by admission of Mozilla itself) is even less likely.
The work you all are doing is incredible. Thank you, and I’ll be a paying customer as long as you keep the show going. I’m notoriously stingy about paying subscriptions for tech services, but the money I pay for Kagi directly impacts my productivity. It feels like a distant nightmare when I think back about the days when I spent time sitting through blogspam results on Google :p
Any chance you would consider adding cryptocurrency payments as an option? You would load up the account with these prepaid credits, and then use whatever plan you want until the credits run out. This is how I pay for Mullvad and works well for me.
I think would be nice as I see kagi is aiming to be privacy-forward. And it’s nice knowing that such a payment is a one-off and won’t forever keep renewing.
Yikes, that thread reflects so badly on Kagi to me. Over a year of people explaining why Monero is needed if you actually cared about user privacy, with no results. Weird claims that it is difficult to implement. Can only conclude that the company doesn't actually care about user privacy
And pray tell, how does one acquire this Monero that you're defending with religious fervor? By linking your credit card to some "currency" exchange that may or may not already be under investigation for fraud?
Even if it were a process as invasive and painful as getting a bank account, it would remain the only technology that enables users to get a search account privately (since Kagi says cash is not scalable for them). Given the level of snark and off-topicness, I presume you don't actually expect me to answer the question with dozens of different methods.
This is a company that claims over and over that they are privacy-respecting, but are running a service that unnecessarily collects your real ID via credit card, otherwise you can't do searches. After seeing what happened to Google, I am shocked that people like you are so dismissive toward me for trying to hold them to account and caring about this issue. Yes I react with religious fervor at this issue - privacy. This is an age where companies are handing over information to convict people for seeking abortions even in the US.
If you have another solution then please share it.
you have to trust not just kagi: the server hosts kagi uses, kagi security policies, all kagi employees, the country kagi operates in, and kagi as a business in perpetuity (not getting bought or drastically changing just to survive).
I can't trust that forever. So it's really just simple that they should have a way to dissociate accounts with searches and payments properly
If you use your credit card with an exchange to buy Monero, then use that Monero (from a non-custodial/self operated wallet) to pay for Kagi search (or Mullvad or anything else) you’re not linking the ID tied to the credit card to your searches and other activity. That’s the point.
There is a lot of value currently tied up in crypto around the world, with increasing difficulty in using it on products and services. Offering a way to use this value would open additional revenue streams for Kagi. The challenge is in conversion, as many banks do not support the practise, and laws are becoming tougher with regards to crypto payments.
One thing that I don't quite get about your family plans: why is the smaller plan ($14 for 1.5k queries) cheaper per query than the larger one ($20 for 2k queries)? Not that it's a big difference, but it feels off somehow.
Product prices are determined by the willingness for consumers to pay for them. Families (especially kids) have different priorities and search habits to IT professionals, for example, which appears to be a common persona for them. It is likely that they bet that the four extra users and child friendly options are what families care about, rather than a high number of searches. It's also likely that families simply don't need that many searches, and in limiting them to 2,000, they discourage abuse (six professionals purchasing one family plan).
> parent controls, which was never a thing in the last 25 years for some reason.
IE tried to setup a rating scheme for websites to adopt for this purpose but it never got any traction. Someone wrote an interesting post about it once but I can’t find it.
Hi. Would there be any chance of you releasing the index of how many trackers websites have publicly? I remember reading about that somewhere on your site, but you eventually took that offline.
I feel that programmers are the main audience that knows life can be better in software. A typical programmer might be so offended by Pinterest's awfulness that they'll look for any way to prevent it, knowing it can be done. Some might go for an extension like uBlacklist, others might be so furious that they want to pay money to stop this somehow.
But seriously, I do think that unless you work with software, it's hard to even imagine alternatives to the big players could exist.
As a European IT professional and a Kagi subscriber, I get my information from and contribute my knowledge to the same sources as you do. I don't know why I would bother to look something up in my native language when there is much more information available in English, the lingua franca of the (Western) world. I don't even read Wikipedia in my own language, unless the subject is better covered.
I’m also European and subscribe to Kagi. One feature I really like is that I can specifically make a local search by adding a bang with the country, like !fr or !es.
Mainly the foxnews.com and breitbart.com being in the top 15 (top 10 if you count all the Pinterest domains as one) in the Block list. As a non-American, I hardly ever see these in my search results, and wouldn't have felt a need for a Block filter for them.
And while there's some British news presence (Guardian and Daily Mail), and what seems to be a Chinese one (kknews.cc), American news media is the one with the heaviest presence (CNN, NY Post, NY Times, etc.).
It shows the demographic of those who would customize their search results. Any search engine that provided this feature would probably skew this way if you looked at who used the feature.
Though since they offer it and others don't, your point still applies.
To me, the results seem interesting regardless of whether they are representative of the population or not. There are questions you can answer without relying on that assumption.
For example: "Can you imagine what your search results would look like if you had access to this tool?"
Mulitply that by the factor of ~50. Not all users use the feature or if they do, block same domains (our user base is very international).
We plan to launch full stats (including number of members and queries/day) next week. "Domains" is meant to be just one of the tabs on this page that got pushed earlier to prod.
I'm surprised Reddit isn't higher on the list. Either I'm on a desktop and Reddit's website is a bit clunky to use - or I'm on mobile and I have to click past several attempts at them to funnel me into their app.
It is a true shame that so much useful information gets trapped within their ecosystem.
A lot of people actually disagree with you—Reddit is number four on the Raise list and on the Pin list, and every few months there's a thread on HN talking about how the only way to get useful information out of Google is to add "reddit" at the end of the search.
Clearly not, they think it's clunky and desktop and too annoying on mobile, (I agree but put up with it often) they're not saying they want more of it.
I was indeed talking about the lower list - simply because of the information presentation is so poor. The information itself is usually quite high quality.
> Either I'm on a desktop and Reddit's website is a bit clunky to use - or I'm on mobile and I have to click past several attempts at them to funnel me into their app.
I have the old-reddit extension on every device I use. It's my 3rd most essential extension, after uBlock Origin and Kill Sticky.
I strongly suspect that if they kill it off at some point we'll have solutions like archive to work around being forced into their UI - even if those solutions are forced to scrap the information out of their web presentation rather than via API calls.
As siblings have pointed out, reddit is actually popular for raising.
For me, what distinguishes it from a stereotypical "block this please" site like Pinterest is that, clunky as the UI may be, content on reddit is often relevant and well linked up with other pages on the Web.
Compare that with Pinterest which I've taken to describe as "the universal sink of the Web" because it seems to get linked to from everywhere but then it's a dead end with zero non-repost content 100% of the time.
Well actually, come to think of it, Pinterest hasn't figured much in my search results ever since I started using kagi a year or so back, and really it didn't get linked to very much from anywhere but the Google results page... Guess I'm getting my money's worth in one more way then I realized. :D
I certainly agree about Pinterest being where data goes to die - and for Reddit I agree that it has an extremely high quality of information, it's just held back by presentation. If I can find the information I'm looking for on a blog I'll usually prefer that but Reddit usually has a good deal of commentary on whatever the topic I'm looking for is.
On reason that Reddit is valuable is that it often contains information that, as far as I can tell, literally doesn't exist anywhere else. This is especially true for product category research. Other than Reddit posts, any product that is even vaguely not mainstream, the only information other than Reddit will be obvious marketing blog spam that doesn't actually help differentiate anything. Generally, I'll check wirecutter, and if Wirecutter doesn't have an article about it, then Reddit is the only other useful option. If it's the only option, the presentation doesn't really matter, although I disagree with you on how bad it is. I still quite like old.reddit on desktop. Although admittedly my mobile experience has been severely degraded since my old preferred app got killed.
> ... or I'm on mobile and I have to click past several attempts at them to funnel me into their app.
It’s not really less inconvenient in terms of time or screen taps in my experience, but changing the subdomain to “old” will at least skip the app prompts. The day that option dies is a day I put a new 0.0.0.0 entry in my hosts file.
It won't help the clunky mobile site much, but if you're on iOS there's a Safari plug in called "Sink It for Reddit" that hides the attempts to funnel you into their app.
I successfully use it to automatically link to Old Reddit. I guess it can also be used to rewrite user-hostile websites' URLs such as YouTube/Twitter/etc to less hostile frontends such as Invidious or Nitter.
This is cool but imo it makes more sense to have URL redirection as a browser extension. That way all twitter links resolve to nitter, not just ones from a search engine. I use this one https://einaregilsson.com/redirector/
Ok wow, seeing these lists might have just made a customer out of me. I’ve been vaguely aware of Kagi and know the general premise but this made it real.
Similarly, I had no idea about this feature, but I now know I want it... I think I just need it to be cheaper, maybe $5 unlimited (or enough for me, which I think is certainly higher than 300, probably higher than 1000, at least without behaviour change) and I'd be in
I assume it's a bit chicken and egg - marginal costs of new customers/extra searches is relatively low, it's the crawling & development/SRE that's expensive - so should (or at least could) come down as more people adopt it, driving further adoption from relative cheapskates like me?
It seems great, but I use DDG 'bangs' (which are also a Kagi feature) loads of times a day, rather than going directly to websites - especially !w & !wikt, !arch & !aur & !archpackages only slightly less - I'd blow through the monthly limit in no time on that alone. (Which also seems a bit unfair, they're just redirects, not using index, there's an argument they shouldn't be included? I suppose if you're so motivated you can work around by using Firefox's feature for it instead of search engine anyway.)
Edit: oh wait, I just found in the FAQ that 'bangs' actually aren't counted as searches. I may have to trial it, see how many searches I actually use how quickly. (And how reasonable the 1.5ç overage seems in my usage.)
Oh really? I just commented about how this misconception made me change my behavior. Wish that was more clear, or that I had higher reading comprehension.
In general, assume that we pass down the savings to the user whenever we can.
Redirecting a bang costs us basically nothing, so it does not count as a search.
Also reloading the same search within a short time (~2 minutes, for example coming back in browser after clicking through a search result) does not count as an additional search as we served cached results.
Another tip: We have decent documentation (that is also open source and editable) and you can access it quickly from Kagi with the !help bang, for example
> In general, assume that we pass down the savings to the user whenever we can.
This is great to hear but it doesn’t really address the point you’re replying to:
> Wish that was more clear
As someone who is just seriously noticing Kagi because of this HN post, I have no brand impression that tells me you’re trying to pass on the savings (most companies don’t) and the marketing material even makes me think you’re targeting a premium price point with healthy margins.
I’m also going to suggest a do not do: Do not make cost savings a public-facing core value. Racing to the bottom usually erodes the value prop while simultaneously opening the door for low-quality competitors.
I would suggest (and caveat: I’m not a marketer, just someone who has seen a lot of brands come and go) either of these directions:
1. Be clear about the cost savings when talking about the specific features. Example: Add a new heading to kagi/features/bangs.html that says “Bangs are Free” and talks about how bangs do not count towards your searches
2. Remove it from being part of your public values altogether. Lean in to the value prop of having a cost and let users self-discover what’s free. If it’s the right choice, highlight it privately: make it a line item on the monthly invoice with something that describes why they are free or put it in onboarding material.
I’m a former DDG user and current paying Kagi customer. One critical difference I’ve noticed is that frequently used bangs with DDG, but almost never with Kagi.
I think this is because over a very short period of usage, I applied block/lower/raise/pin to various domains, and now my search results are almost always perfect.
That said, I understand the value of going directly to a site. On that front, Kagi makes it trivial to define new bangs.
Lately I’ve moved to doing this at the browser level instead of via a search engine. It’s not a huge speed difference, but in principle it seems wrong to have an extra round-trip out to a search engine to redirect me to another site, when I can set up the keywords locally and go directly. Firefox makes it easy to do: https://support.mozilla.org/en-US/kb/how-search-from-address...
I've used DDG for years before switching over to Kagi and I was accustomed to having to use the !s bang daily (and !g before I found out about !s). On Kagi I literally never use it, since it already incorporates Google search results and filters out a lot of crap.
I've also defined some custom bangs to search private Jira and Confluence instances with some default filters applied, because the default search on those instances was driving me insane. It took me a long time to convince myself to pay $10 a month for a search engine, but I've never looked back.
I'm currently experimenting with hostname rewrite rules.
I was worried about the 1000 search limit, and it has so far (since beta) proven not to be a problem. I use Kagi for all my searches on all machines and my phone and average 800 or so a month. I haven’t modified my behavior at all, and consider myself a heavy search engine user.
I have a theory that I would actually need to do twice as many searches with Google or DuckDuckGo or whatever, since the SEO spam would force me to do more term refinement. With far less of that (and a tiny tiny bit of settings tuning) I get better results much quicker. I’d test it but I have a job to do and can’t swallow the idea of going back to how bad the other viable options really are.
That sounds appealing, but there are so many $10 and $15 monthly warts on my balance that the burden of adding another (with even a minor overcharge fee) feels pretty high. I wish their lower tiers had double the current limits.
> I have a theory that I would actually need to do twice as many searches with Google or DuckDuckGo or whatever
FWIW, I counted searches I made with DDG (by parsing my FF history export) before joining the beta, and it was slightly over 1k, with Kagi my searches are in the 700-800 range.
> I have a theory that I would actually need to do twice as many searches with Google or DuckDuckGo or whatever
This is a great point that should really be highlighted more in their marketing. It was obvious once I read it and shows two obvious benefits: 1. There is probably zero issue with fitting in to the 1000 searches per month. 2. It saves time for many of the searches we’re already doing.
> 1. There is probably zero issue with fitting in to the 1000 searches per month.
You are never limited to 1000 monthly searches, this is a great misunderstanding about Kagi. After passing your threshold, you are charged 1.5c per query.
I think that’s clear enough on the details of the plans. What I meant is fitting in without incurring extra charges (no point in arguing that 1.5c is small, that’s not the point)
Defaulting to DuckDuckGo and using a bang for kagi when I don’t immediately get what I want would be nice. Unfortunately there isn’t one. I guess I could use an extension that lets me add personal ones.
I didn't say it was a 'killer feature', I said 'now I know I want it'. There are very many things I want but not at the price they cost!
I don't think coffee's a great analogy - it has far more utility for me, I don't buy it as a service, and it costs me roughly £24 (what, $30ish) a month in beans.
I'm not seriously suggesting it because I think comparing prices of different things to determine value very quickly gets silly, but a closer comparison might be Netflix: Kagi would cost me about as much per month for search as Netflix. Silly, as I said, but if I tried really strictly to pay for things according to relative value or utility to me, there's no way that would make sense.
Part of the problem is asking me to switch from free I suppose - if DDG suddenly started charging me $5 I'd be more likely to pay Kagi $10 than I am today if that makes sense.
People always compare with Netflix, but that makes no sense at all to me. Is Kagi comparable to Netflix because they are both presented to you on a digital screen? I see nothing else that makes them comparable.
Laundry detergent and breakfast cereal both come in a box in the supermarket, but it doesn't make the products comparable. A house and a car both have a door and you usually sit inside, but it doesn't make the products comparable.
Netflix is a great bargain (for those who enjoy their offerings), and I think people are shooting themselves in the foot by dismissing other great services because they've somehow tricked themselves to thinking they should be compared to Netflix only because they're offered through a digital screen.
I'm at the $10/mo level, and am totally happy to keep paying. I rely on search throughout the day for my job, and for my personal interests. $10/mo for an ad-free tunable search that provides good results is a completely worth it. And, I'm more than happy to vote with my money to support the web that I want to exist.
Otoh (for contrast) I run adblock and sponsorblock on youtube, and if they start blocking traffic from people like me I'm going to cut back my YT viewership significantly (and maybe try harder to look for stuff on nebula instead, which I never browse). While I do get some educational value from YT, it's limited, and I guess I just fundamentally object to the ad-first business model.
You should install the Kagi browser extension and try out the Summarizer functionality on YouTube videos. I’ve found it pretty useful for videos that are needlessly filled with fluff.
I subscribe to the cheapest plan, and only use Kagi for non-trivial searches. That’s more than enough, I don’t need a paid search engine when I just type a website’s name instead of its url.
I set it up with DDG/Google as default, and for Kagi I start with “k “.
I have gone over the limit (on the grandfathered 1500 search plan), and paid the extra $1-2 for it. It's fine.
The cost of search has made me think about what terms I use now and what I search for. Google made me lazy, I now use my phones world clock app rather that searching for "current time X", and word my searches more thoughtfully.
It has blown my mind for years the extent to which the Google Image Search team let Pinterest completely undermine the usefulness of their product, and they just kept on giving pinterest the top search result spots.
It's honestly more useless than the pages with a list of 1000 phone numbers that you get when you try to search for who an unknown number is.
Does anyone even work on Google Image Search? Is it run by a bunch of undercover Pinterest employees who infiltrated the team and pushed out the Googlers?
I'm honestly surprised by all the Pinterest hate because I don't think I've ever once, not a single time, left clicked a Google Image Search result. It's like drinking from the dead. You just don't do it.
>I don't think I've ever once, not a single time, left clicked a Google Image Search result. It's like drinking from the dead. You just don't do it.
You just don't do it according to whom? I would consider viewing the fullsize image to be core functionality for an image search, and it used to work just fine on GIS up until a few years ago.
Umm, according to the latest data, user choice is far less competitive than algorithms at stabilizing engagement metrics. Users' behavior speaks louder than what they believe in their little heads.
my little head sometimes has me patch and spin up old pc's which should be a fairly meaningless and pointless exercise if it wasn't for those marvelous algorithms. Each box represents a mental time capsule worth of youtube suggestions. It attempts to lure the viewer into popular shit everyone else is watching but if you carefully avoid those thumbs and click only on the things seeded initially it will gradually show more and more of it. Things like small minded peasants farming their little farm by hand in languages completely unknown to me. The next box is seeded with "full movie" with an algorithm trying to carve out my taste. I also have A box full of little prepper people.
On my main pc it is 100% convinced I want to watch starcraft, technology, coding and fitness videos. Im not completely uninterested but the goal of the design seems to make my mind even smaller. It is constantly signaling that... how put it.... it thinks Im shallow as fuck.
Then again on the other box it is convinced I want to see nothing but old people planting rice. Its probably not good for its world view. If it was conscious it would be a cruel thing to do.
Or perhaps an amoral and unrestrained trillion-dollar paperclip-maximizer is, within certain paramaters, more effective at steering monkey-brains than said brains' resident consciousnesses are.
Definitely the best feature of Kagi! I sometimes wish there was some way to slightly adjust how this works. As I also have Wikipedia on raised priority, I'd still like some official product/company/etc. website to rank higher than the Wikipedia article of it.
Also, a custom priority ranking among the raised importance domains would be nice.
Dearrow. An open source and crowdsourced tool replace youtubes stupid fucking clickbait title cards, and replaces many titles with better ones. You can contribute!
Pin means that if a result from that domain would appear anywhere in 50 or so relevant results that Kagi usually surfaces on the first page, it would be 'pinned' to the top of results.
I guess you're right. I haven't seen them in a while either. But for a while there they would be among the top results for all kinds of searches, and the content itself was always garbage.
But I don't get pinterest very often lately either. Not never but not most of the time either.
I found the right wing media blocks interesting. I wondered if that was in response to getting search spammed by them or some "principal". Because I don't think I've ever seen Fox News or Breitbart in search results anyway.
Same comment for CNN and wapo that I see are also there. Maybe I'd dont search current events enough, they all seem like odd things to block to me.
Fox News is the same level of journalism as guardian, verge and nytimes. As well as Wapo and CNN. I guess people are choosing to be in a bubble of the trash they like. Maybe because they don't know the billionaires that own the other media they think they are decent news sources.
Other surprising trash to me is people want to see: rtings.com and goodreads.com.
I really tried hard with Goodreads and never found good book recommendations there.
And the NYT is publicly traded (The CEO has an estimated networth of only 10mil)
Verge doesn't really do journalism that makes me worried about who owns it.
> Other surprising trash to me is people want to see: rtings.com and goodreads.com.
rtings does colour profiles of plenty of slightly less popular monitors that no one else does. It's been an invaluable resource for me.
This comment is genuinely nonsensical. I'd think that it's obvious that rich owners tend to push a conservative economic spin to preserve their wealth.
> I really tried hard with Goodreads and never found good book recommendations there.
Most people end up on goodreads from search to see the reviews people left for the book....
I actually did a smaller thing for HN. I have client side JS that blocks results from a few domains that seemed to promote low quality discussions and/or irrelevant submissions.
shrugs I don't really feel the need to defend the contents of that list. If the resulting conversations suck, aren't relevant to me, and clog up the HN front page - in they go.
There's something uniquely social-media-y about a person posting a list of websites they don't enjoy the contents of or discussion around, and then posting that they've filtered them out in a way that only impacts them, and then that person getting questioned and challenged on their filtering.
> Off-Topic: Most stories about politics, or crime, or sports, or celebrities, unless they're evidence of some interesting new phenomenon. Videos of pratfalls or disasters, or cute animal pictures. If they'd cover it on TV news, it's probably off-topic.
The domains in question lead to guaranteed uninteresting and routine political bickering where each side is compelled to recite their talking points. It has nothing to do with the reporting and everything to do with how people comment on those topics. I'd be guilty of it too if I weren't busy flagging those posts.
Isn't VOA (and BBC for that matter) government owned & funded? That's enough right there for me to doubt their "reliably factual" reporting. It's hardly a political opinion- the owners of these services expect to get some return on their investment. Because they're not in it for the money (they can just create more money basically at will), then why are they in it? Likely to help craft a narrative that suits them.
At least for the bbc, the government isn't directly involved in the editorial. They have to resort to appointing their friends director general & threatening to abolish the licence fee to try to get what they want.
VOA and BBC are not alike at all in my opinion. Many countries in Western/Northern Europe have government funded sites run like the BBC and they are mostly in the top of unbiased and free journalism. They also create a lot of high quality documentaries and crime thrillers.
Nothing is perfect and of course there's a certain kind of culture on each of these workplaces that will mske it either more left or right up or down, but they aren't controlled by politicians, just like a hospital isn't told how to treat patients by politicians.
I'd rank BBC News above all US news outlets (and I'll just add I'm neither from the US nor UK).
Could this be modified to block comments from a list of usernames too? The last time I looked at HN’s DOM made me sad.
There are some prolific users who… while their comments are of normal quality, trigger a strong desire for me to respond. I’d rather simply not see their comments (and reply trees) at all.
Personally, I never reply. I only lurk. Because every time I reply I get sucked into the social media nightmare that is waiting for a response to my reply. So I opt out of that by never, ever replying, no matter how big the urge is to reply. I never do it. Almost never. 99.99% of the time, I skip replying. I think you might lack a discipline.
Comments Owl for Hacker News [1] adds a mute control to comments and user profiles which lets you do this.
If you want to roll a quick version of your own, once you've identified rows containing comments you want to block, you need to hide all subsequent rows which have a higher indent. I see there's now an "indent" attribute in the DOM which would make this even easier.
I've never understood the hate for w3schools. No, it's not MDN, but it's not offensively bad either, and it's been a helpful reference from time to time. If w3schools comes up first in a search, I trust that it probably has the answer to whatever very simple question I must have asked, unlike something like Pinterest, which will just be spam no matter what.
It was offensively bad and incorrect, and while I've heard it has improved since I think a lot of people just remember how it used to be and don't use it.
They had functional but inaccurate articles (in terms of best practices and long term solutions). I avoided them for a time but I must admit they listened and amended their tutorials, which is more than a lot of websites can say.
Historically, at least, it had a lot of examples that technically did what they were supposed to do, but in ways that you really, really, really shouldn't be doing them.
I knew many developers that used w3schools that thought it was official source because w3c, I know it's not completely the sites fault but they named the domain. That's one reason I dislike, additional the quality of w3schools is compared with mdn for beginners that are not interested into the details.
It's more a historical thing. Over a decade ago they were always on top, with often wrong/misleading/bad practice/insecure/simplified stuff. And lots of people thought of them as some kinde official entity because of their name similarity with W3C / w3.org
It's much better now, though. But for lots of old school frontenders the reputation stuck.
In your opinion. W3fools only pulled an “it’s ok now bro, don’t sue us” update. Anytime I give it a go, I’m still presented with inaccurate and harmfully incomplete results.
Want to learn from a “for dummies” website? Be my guest. But my hate is fully justified.
w3schools is only bad in comparison with MDN, but it still always rises to the top. I don’t think people have a problem with w3schools as such, but with the fact it always bumps MDN.
W3schools intentionally placed themselves as a representation of the W3C (similar to what Fedex does with its name) and sold, still does I'm sure, certificates under that misconception. These were worthless. This is in addition to the other complaints.
501 comments
[ 3.0 ms ] story [ 337 ms ] threadAs far as I know, this is more a failure of Google's inexplicable algorithm than Pinterest doing anything specifically bad.
Guess it depends on whether it's UGC entirely or whether the CMS could note these kind of pages. Again, shouldn't be hard if it's just a simple link as the main content of the page.
Google are the ones who should be looking at the situation and asking if this is the state they really want.
//added EHow is an example of a site that was flagged
My most used page is this one when I need to jog my memory: https://www.w3schools.com/cssref/css_selectors.php
It's much better than this one: https://developer.mozilla.org/en-US/docs/Web/CSS/CSS_selecto...
Or any of the other results on the first page
w3schools tend to oversimplify the concepts to the point of technical inaccuracy.
and the tone of the articles read more like tutorials instead of reference manuals.
I'm sure you meant w3schools but w3.org is where a lot of the official documentation for web technologies is and not affiliated with w3schools.
https://web.archive.org/web/20110412103745/http://w3fools.co...
This site [0] used to be the hub for the hate and now has a message saying that w3schools has gotten better. If you're curious, they still have a link to the web archive that has all the mistakes documented.
The most egregious of their mistakes to me:
> […] professional web developers often prefer HTML editors like FrontPage or Dreamweaver, instead of writing plain text.
[0] https://www.w3fools.com/
1. The ads make the fan on my laptop spin (I prefer to avoid ad-heavy websites over using adblock)
2. The 'Try it yourself' button is fake. It doesn't actually run the code and doesn't let you edit it.
3. It is not a good reference. To give an example, the last time I visited w3schools was upon searching "react ul li", which landed me on the page https://www.w3schools.com/react/react_lists.asp . Ironically, the reason why I was searching the term, was to find advice on setting the "key" prop in on list items. The w3schools example doesn't even mention the prop and produces errors. Compare that to the official reference which ranks much lower in Google search: https://react.dev/learn/rendering-lists#where-to-get-your-ke... .
All that to say, I think beginner oriented well SEO'd content is going to be polarizing, which makes Kagi's approach great for everyone, if you want it it's there, if it annoys you, quickly block it
I wish I hadn't. There were many innacuracies, outdated advice, and over simplifications.
I came up on a slew of weird and inconsistent sources, and that bit me when discussing things in college. I often felt over confident because of these type of sites.
W3schools (used to at least) gave a very shallow intro to programming languages (PHP, ASP, etc), and I remember a professor basically laughing at me once because of that misinformation.
So, maybe others had experiences like mine.
EDIT
They also had these weird "certifications" for technologies that were basically really dumbed down quizzes. They were made to make people feel special, I think. Dunno, I did them all, and I almost got laughed out of class.
To be fair, it is biased, but then so are most outlets one way or the other: https://www.allsides.com/media-bias/ratings
(This isn't to say that other major media outlets don't have problems with accuracy. But the magnitude is very different.)
Also, as an aside. What is a "false claim"? If it's a claim, they're asking you to believe it, which means they need to motivate it and provide details to back it up. Seems precisely where all the media generally is (including CNN/BBC/etc), pandering to peoples' bias without saying anything factually incorrect.
Watching Glenn Greenwald and Matt Taibbi repeatedly roast NYT, CNN and MSNBC would make that obvious to anyone paying attention.
The magnitude is different indeed, see the Russiagate hoax. The whole entire thing was made up. And it wasn't Breitbart the one perpatrating it.
They primarily used "Experts Exchange" and http://www.experts-exchange.com/
If you think through the feature, it's not at all obvious that it's a good idea for an established search engine targeting the general public [0]. But even then it might be a great idea for a search engine that's trying to establish an initial niche. It doesn't even matter whether the idea actually works well in practice: if 100k people think that manual curation of search result domains is a high value feature and are willing to pay for it, the feature has done its job.
[0] I wrote a more detailed comment on why a few months ago: https://news.ycombinator.com/item?id=34294800
Which suggests to me that while kagi might be good now, if it gets popular it may undergo enshittification in an effort to monetise its popularity.
For Kagi, it's a matter of when, not if. Capitalism guarantees it.
Both are technically possible, but so extremely unlikely, that you can rely on them not happening.
I concede that if they got big enough, someone would come with enough money that they couldn't refuse, and buy them to try and "unlock vale" by racing to the bottom.
Don't confuse an entire economic system with the unique problems caused by public ownership or investing (i.e. growth at all costs). Plenty of smaller organizations would be quite happy to operate with modest profits for their owners and never scale beyond that point.
Precisely, which is why enshittification is a when, not an if. A paid search engine will never pick up enough users to survive at $10/mo. It will either be acquired or start throwing in "non-intrusive" ads to balance the books.
Here is a personal guarantee that this will never happen with Kagi while I am in charge.
There are too many ad-supported search engines and browsers out there. I would not waste 10 years of my life to make yet another one. If Kagi can not sustain iteself with memberships, it will be the end of it.
The initial reason I started Kagi was to provide my kids with an opportunity to grow up in an ad-free experience of the web. In this interview [1] I talk more about my motivation.
[1] https://dkb.io/post/DEPR_kagi-interview
Call us back in 6 months when you arbitrarily change something else and give us another motivating pep talk about your commitments. roll eyes.
I was just about to finish this post and ask if anyone knows the history. I Googled that and was surprised by the answer from Wikionary: https://en.wiktionary.org/wiki/enshittification
God damn, "Dr." Doctorow again!Kagi has already put itself on a path where it doesn't have to make itself worse to cover its expenses, because it is already covering them.
Imagine if one of your friends come back from a holiday trip and tells you how fantastic the place they visited was. Is your response "Well, maybe that place ain't going to be so nice in a few years!"?
The usefulness of a per-user block list is not fighting outright spam. The statistical filters, etc, as applied by the search engine, work best here, as you correctly state.
The point is to remove certain well-known sites with information about well-known topics, which is unhelpful for the particular user, me. That information is technically not spam, but it's on a wrong level, badly presented, etc.
That is, if I search some info about a particular CSS trick, I want MDN be on top, some professional blogs be near top, and w3schools be on the third page. For someone else, priorities may be different, but I know what to expect for the search queries I issue, and I'd be glad if the search engine helped me.
Equally, certain domains are known to be more rich with information I happen to seek, like reddit.com, so I'd like to give them a boost. Some other people may prefer something else, or not give any boost at all.
So, this is personalization, not anti-spam. If Google kept it, it could even bring important insights into what ads I would skip and what would click :)
Maybe 90% of users would ignore this feature. But the 10% of power users maybe disproportionately important: with more disposable income, more clout on social networks, more decision-making power. I suppose that the shutdown of Google Reader, which affected a relatively small number of users, affected all the wrong people, and destroyed so much goodwill that the losses from it are many orders of magnitude larger than whatever savings from not supporting it any more.
I'd count that kind of low quality site as spam, and the problem is that there's a basically an infinite supply of them. The boost feature is more interesting in that the curation is at least not a never-ending task. New high quality sites that I know about don't pop up all that often!
It's still not an entirely robust feature since you'll need to do a lot of quality work to prevent mediocre results from a boosted site from shadowing better results from unboosted sites, especially for users with a ton of boosts. But that quality work is much easier to justify when the manual curation is one of your core selling points and used by a large proportion of the customers vs. when it's a feature in a dusty context menu that basically none of the userbase interacts with.
> Maybe 90% of users would ignore this feature. But the 10% of power users maybe disproportionately important
I'd be shocked if the number of users for this kind of feature was anywhere near 10% of all search engine users. Something like 0.1% seems much more likely. But again, if those 0.1% really want that feature it's a great fit for a power user search engine.
It still doesn't mean that it's a feature every search engine should obviously have.
That strikes me as a non-goal. As the search engine, you can't know what attributes a user is valuing when they boost, and so you can't know when it's appropriate to override that boost. "Better" is subjective, contextual, and comes in thousands of different flavors.
But when they do, how are you going to know if all you see is the mediocre spam instead? No, I want to block the domains which are never going to be a good result. This may actually surface things I will want to boost.
NYTimes is another good example, great content for subscribers, essentially useless spam in a search result for non-subscribers. The only person who can configure this is the user.
Is their information of such low quality ?
They’re better now but this is timeless.
I don't think he did. He did say " it's not at all obvious that it's a good idea for an established search engine targeting the general public"
And looking at the data from the link, other than wikipedia and reddit, all the pinned domains are for developers or technical folk.
So it looks to me that kagi is not popular to the general public. If it was, programming-related sites wouldn't so heavily top the popular list.
When you’re first learning a subject you often want simple lies more than complicated truths.
Like pinterest.
Goddamn pinterest.
Just the amount of times i've used "-pinterest" in search is staggering.
Now, I'm not blind to the fact that Reddit's user demographic has a lot in common with us HN folks, while Pinterest... well, not so much. It leans towards a non-techy crowd, with more women users. But the way I see it, if Google keeps popping Pinterest up in search results, someone out there must be finding it helpful.
Perhaps what this all boils down to is a bit of a demographic reality check, rather than a definitive statement on any one platform. And just to be clear, I'm not standing on a soapbox championing for Pinterest, but I reckon we could probably nudge Google to get a bit smarter with its algorithms.
So, what if instead of giving domains the chop, we pushed for Google to up its game on the personalization front? Like, if I'm consistently booting Pinterest from my searches with "-pinterest", Google could take the hint and drop it down the ranks for me. Wouldn't that make more sense than completely blacklisting sites that some folks might still want to use? In the end, we all win if Google gets better at tuning into our search habits, right?
Same for youtube, a simple blacklist would make my search results much nicer, but youtube seems to push a few large media companies as hard as possible.
> Third, some people will probably block domains they shouldn't have blocked, and then have a bad user experience in future searches as the sites with genuinely best results is blocked. And then you're only left with only bad options: ignoring the users' stated preferences which they'll hate, or serving bad results that they'll also hate.
That's a real possibility, but it can be addressed by presenting a block that would say, for example "results from blocked domains" or some such.
This is bad for Google's bottom-line thus why such a feature was removed and will never be coming back.
1. Sell the first few results but make it look organic (this all happened behind closed doors and I knew a few people who paid for such results)
2. Hierarchical/topic based collections of links (This is what Yahoo was famous for in the 90s if I remember correctly, I'll keep saying that because it was a long time ago!)
The whole "google cares about you" was an outcropping of the organic flocking to google because of their reliable and algorithmic results that just seemed trustworthy, they capitalized on that and it honestly lasted a lot longer than I expected before they really started being evil.
https://iorate.github.io/ublacklist/docs
Google didn't prioritize it's visibility [intentioally] over the past 2 decades as well. Maybe it's time to do so.
I can't imagine how tedious the process may be to make very small changes on the main google home page.
The features you were referring to were removed a few years after they were launched.
https://github.com/quenhus/uBlock-Origin-dev-filter
I wonder if they are interested in attracting users with more diverse searching requirements, and if so, how they would be able to actually reach them.
I would say this is a blessing as it allows us to focus on things that matter to tech-savvy users (Kagi founder here).
> I wonder if they are interested in attracting users with more diverse searching requirements,
Yes, we are :)
> if so, how they would be able to actually reach them
Through our family plan [1], most likely. This is how I got my family members on board a paid search product.
I might add that "Kagi for Kids" is the first true attempt at having a search engine made specifically for kids, with a host of parent controls, which was never a thing in the last 25 years for some reason.
We are also seeing more and more users who are not techies (but are tech-savvy) becoming members. This may be followed by non-tech savvy users in the future. I do see the first signs of the age of ad-supported, "free" search, coming to an end [2].
[1] https://blog.kagi.com/family-plan
[2] https://blog.kagi.com/age-pagerank-over
Unfortunately we can not influence the choices that Safari makes, but we did pull an effort to build an entire browser [1] to replace it because of that :)
[1] https://browser.kagi.com
Thank you. Lots of love to Kagi, though I'm not a paying user yet but I do love what you folks are doing and want to f*ck Google.
The macOS browser is based on WebKit. If the team can’t justify the Linux version yet, changing it to a whole different engine (and one which is notoriously hard to embed, by admission of Mozilla itself) is even less likely.
[1] xSearch for Safari: https://apps.apple.com/us/app/xsearch-for-safari/id157990206...
Any chance you would consider adding cryptocurrency payments as an option? You would load up the account with these prepaid credits, and then use whatever plan you want until the credits run out. This is how I pay for Mullvad and works well for me.
I think would be nice as I see kagi is aiming to be privacy-forward. And it’s nice knowing that such a payment is a one-off and won’t forever keep renewing.
[1] https://kagifeedback.org/d/493-enable-anonymous-payments-ala...
This is a company that claims over and over that they are privacy-respecting, but are running a service that unnecessarily collects your real ID via credit card, otherwise you can't do searches. After seeing what happened to Google, I am shocked that people like you are so dismissive toward me for trying to hold them to account and caring about this issue. Yes I react with religious fervor at this issue - privacy. This is an age where companies are handing over information to convict people for seeking abortions even in the US.
If you have another solution then please share it.
Dozens of different methods to ... buy crypto? You can keep them.
I would trust something like Kagi more than any crypto exchange to keep my privacy.
You are right that it may not be a good idea to have all your searches tied to your identity, but mixing crypto into that won't help.
I can't trust that forever. So it's really just simple that they should have a way to dissociate accounts with searches and payments properly
I’ve seen quite a few marketing people and researchers in (their discord channel) #introduce-yourself
IE tried to setup a rating scheme for websites to adopt for this purpose but it never got any traction. Someone wrote an interesting post about it once but I can’t find it.
Would you mind posting this as a feature suggestion on kagifeedback.org ? (this is how most ideas for Kagi come to life)
But seriously, I do think that unless you work with software, it's hard to even imagine alternatives to the big players could exist.
And specifically, American programmers and IT professionals.
And while there's some British news presence (Guardian and Daily Mail), and what seems to be a Chinese one (kknews.cc), American news media is the one with the heaviest presence (CNN, NY Post, NY Times, etc.).
Though since they offer it and others don't, your point still applies.
For example: "Can you imagine what your search results would look like if you had access to this tool?"
We plan to launch full stats (including number of members and queries/day) next week. "Domains" is meant to be just one of the tabs on this page that got pushed earlier to prod.
It is a true shame that so much useful information gets trapped within their ecosystem.
It's on two lists.
> Either I'm on a desktop and Reddit's website is a bit clunky to use - or I'm on mobile and I have to click past several attempts at them to funnel me into their app.
I have the old-reddit extension on every device I use. It's my 3rd most essential extension, after uBlock Origin and Kill Sticky.
For me, what distinguishes it from a stereotypical "block this please" site like Pinterest is that, clunky as the UI may be, content on reddit is often relevant and well linked up with other pages on the Web.
Compare that with Pinterest which I've taken to describe as "the universal sink of the Web" because it seems to get linked to from everywhere but then it's a dead end with zero non-repost content 100% of the time.
Well actually, come to think of it, Pinterest hasn't figured much in my search results ever since I started using kagi a year or so back, and really it didn't get linked to very much from anywhere but the Google results page... Guess I'm getting my money's worth in one more way then I realized. :D
Kagi is so worth the money!
It’s not really less inconvenient in terms of time or screen taps in my experience, but changing the subdomain to “old” will at least skip the app prompts. The day that option dies is a day I put a new 0.0.0.0 entry in my hosts file.
I successfully use it to automatically link to Old Reddit. I guess it can also be used to rewrite user-hostile websites' URLs such as YouTube/Twitter/etc to less hostile frontends such as Invidious or Nitter.
I assume it's a bit chicken and egg - marginal costs of new customers/extra searches is relatively low, it's the crawling & development/SRE that's expensive - so should (or at least could) come down as more people adopt it, driving further adoption from relative cheapskates like me?
It seems great, but I use DDG 'bangs' (which are also a Kagi feature) loads of times a day, rather than going directly to websites - especially !w & !wikt, !arch & !aur & !archpackages only slightly less - I'd blow through the monthly limit in no time on that alone. (Which also seems a bit unfair, they're just redirects, not using index, there's an argument they shouldn't be included? I suppose if you're so motivated you can work around by using Firefox's feature for it instead of search engine anyway.)
Edit: oh wait, I just found in the FAQ that 'bangs' actually aren't counted as searches. I may have to trial it, see how many searches I actually use how quickly. (And how reasonable the 1.5ç overage seems in my usage.)
more info: https://help.kagi.com/kagi/getting-started/faqs.html#how-are...
Redirecting a bang costs us basically nothing, so it does not count as a search.
Also reloading the same search within a short time (~2 minutes, for example coming back in browser after clicking through a search result) does not count as an additional search as we served cached results.
Another tip: We have decent documentation (that is also open source and editable) and you can access it quickly from Kagi with the !help bang, for example
!help how are searches counted
will land you on
https://help.kagi.com/kagi/plans/plan-types.html#how-searche...
This is great to hear but it doesn’t really address the point you’re replying to:
> Wish that was more clear
As someone who is just seriously noticing Kagi because of this HN post, I have no brand impression that tells me you’re trying to pass on the savings (most companies don’t) and the marketing material even makes me think you’re targeting a premium price point with healthy margins.
I would suggest (and caveat: I’m not a marketer, just someone who has seen a lot of brands come and go) either of these directions:
1. Be clear about the cost savings when talking about the specific features. Example: Add a new heading to kagi/features/bangs.html that says “Bangs are Free” and talks about how bangs do not count towards your searches
2. Remove it from being part of your public values altogether. Lean in to the value prop of having a cost and let users self-discover what’s free. If it’s the right choice, highlight it privately: make it a line item on the monthly invoice with something that describes why they are free or put it in onboarding material.
I think this is because over a very short period of usage, I applied block/lower/raise/pin to various domains, and now my search results are almost always perfect.
That said, I understand the value of going directly to a site. On that front, Kagi makes it trivial to define new bangs.
Lately I’ve moved to doing this at the browser level instead of via a search engine. It’s not a huge speed difference, but in principle it seems wrong to have an extra round-trip out to a search engine to redirect me to another site, when I can set up the keywords locally and go directly. Firefox makes it easy to do: https://support.mozilla.org/en-US/kb/how-search-from-address...
I use bangs all the time and wouldn’t even use Kagi without them. It’s so convenient when I know I want results from a specific page.
I've also defined some custom bangs to search private Jira and Confluence instances with some default filters applied, because the default search on those instances was driving me insane. It took me a long time to convince myself to pay $10 a month for a search engine, but I've never looked back.
I'm currently experimenting with hostname rewrite rules.
I have a theory that I would actually need to do twice as many searches with Google or DuckDuckGo or whatever, since the SEO spam would force me to do more term refinement. With far less of that (and a tiny tiny bit of settings tuning) I get better results much quicker. I’d test it but I have a job to do and can’t swallow the idea of going back to how bad the other viable options really are.
FWIW, I counted searches I made with DDG (by parsing my FF history export) before joining the beta, and it was slightly over 1k, with Kagi my searches are in the 700-800 range.
This is a great point that should really be highlighted more in their marketing. It was obvious once I read it and shows two obvious benefits: 1. There is probably zero issue with fitting in to the 1000 searches per month. 2. It saves time for many of the searches we’re already doing.
You are never limited to 1000 monthly searches, this is a great misunderstanding about Kagi. After passing your threshold, you are charged 1.5c per query.
I can't remember what for but I'm pretty sure I've used it and gotten a pretty fast approval.
I don't think coffee's a great analogy - it has far more utility for me, I don't buy it as a service, and it costs me roughly £24 (what, $30ish) a month in beans.
I'm not seriously suggesting it because I think comparing prices of different things to determine value very quickly gets silly, but a closer comparison might be Netflix: Kagi would cost me about as much per month for search as Netflix. Silly, as I said, but if I tried really strictly to pay for things according to relative value or utility to me, there's no way that would make sense.
Part of the problem is asking me to switch from free I suppose - if DDG suddenly started charging me $5 I'd be more likely to pay Kagi $10 than I am today if that makes sense.
Laundry detergent and breakfast cereal both come in a box in the supermarket, but it doesn't make the products comparable. A house and a car both have a door and you usually sit inside, but it doesn't make the products comparable.
Netflix is a great bargain (for those who enjoy their offerings), and I think people are shooting themselves in the foot by dismissing other great services because they've somehow tricked themselves to thinking they should be compared to Netflix only because they're offered through a digital screen.
Though I'd stand by Netflix as a better comparison to Kagi than coffee (as in the comment I was replying to) to anyone who insists on doing it.
Otoh (for contrast) I run adblock and sponsorblock on youtube, and if they start blocking traffic from people like me I'm going to cut back my YT viewership significantly (and maybe try harder to look for stuff on nebula instead, which I never browse). While I do get some educational value from YT, it's limited, and I guess I just fundamentally object to the ad-first business model.
I set it up with DDG/Google as default, and for Kagi I start with “k “.
The cost of search has made me think about what terms I use now and what I search for. Google made me lazy, I now use my phones world clock app rather that searching for "current time X", and word my searches more thoughtfully.
Me: Ok, please block these domains: "pinterest.com, w3schools.com, ..."
Google: Sorry, can't do that!
It's honestly more useless than the pages with a list of 1000 phone numbers that you get when you try to search for who an unknown number is.
Does anyone even work on Google Image Search? Is it run by a bunch of undercover Pinterest employees who infiltrated the team and pushed out the Googlers?
You just don't do it according to whom? I would consider viewing the fullsize image to be core functionality for an image search, and it used to work just fine on GIS up until a few years ago.
I get that this must be something bad, but what does it mean, exactly? Drinking blood from a corpse?
I want least engagement with the search engine; it's a fail if I'm spending time just looking at search page and not the results.
On my main pc it is 100% convinced I want to watch starcraft, technology, coding and fitness videos. Im not completely uninterested but the goal of the design seems to make my mind even smaller. It is constantly signaling that... how put it.... it thinks Im shallow as fuck.
Then again on the other box it is convinced I want to see nothing but old people planting rice. Its probably not good for its world view. If it was conscious it would be a cruel thing to do.
oh the thoughts going on in my small mind...
This completely frames it for me.
I was very hesitant to pay for a search engine at first, but Kagi has turned into the “stickiest” monthly subscription I have.
“I am Phil Wang! There’s nothing on Netflix more relevant to me than an hour of my own thoughts!”
https://podtail.com/en/podcast/netflix-is-a-daily-joke/phil-...
https://twitter.com/missdianemorgan/status/16718830928004464...
The average human out there is not boosting developer.mozilla.org
Still...I'm thankful that they publish the stats. It's a useful mental reference point if nothing else
https://github.com/iorate/ublacklist
And here's two bonus things you need if you make use of youtube
Sponsorblock. As it says on the tin, automatically skips sponsor sections.
https://github.com/ajayyy/SponsorBlock
Dearrow. An open source and crowdsourced tool replace youtubes stupid fucking clickbait title cards, and replaces many titles with better ones. You can contribute!
https://github.com/ajayyy/DeArrow
https://addons.mozilla.org/en-US/firefox/addon/sponsorblock/
If so, I would have expected a much longer tail.
But I don't get pinterest very often lately either. Not never but not most of the time either.
Same comment for CNN and wapo that I see are also there. Maybe I'd dont search current events enough, they all seem like odd things to block to me.
Other surprising trash to me is people want to see: rtings.com and goodreads.com.
I really tried hard with Goodreads and never found good book recommendations there.
https://en.wikipedia.org/wiki/Fox_News_controversies
Really?
> I guess people are choosing to be in a bubble of the trash they like. Maybe because they don't know the billionaires that own the other media
https://www.theguardian.com/gnm-press-office/2017/aug/28/the...
And the NYT is publicly traded (The CEO has an estimated networth of only 10mil)
Verge doesn't really do journalism that makes me worried about who owns it.
> Other surprising trash to me is people want to see: rtings.com and goodreads.com.
rtings does colour profiles of plenty of slightly less popular monitors that no one else does. It's been an invaluable resource for me.
This comment is genuinely nonsensical. I'd think that it's obvious that rich owners tend to push a conservative economic spin to preserve their wealth.
> I really tried hard with Goodreads and never found good book recommendations there.
Most people end up on goodreads from search to see the reviews people left for the book....
You were quite willing to promote the specific and lengthy list of disliked-sites to strangers.
A sudden switch to fatigue and disinterest does not seem genuine to me.
>... that seemed to promote low quality discussions
My guess is that these domains are responsible for hot button topics that result in 300+ comment threads.
The domains in question lead to guaranteed uninteresting and routine political bickering where each side is compelled to recite their talking points. It has nothing to do with the reporting and everything to do with how people comment on those topics. I'd be guilty of it too if I weren't busy flagging those posts.
Nothing is perfect and of course there's a certain kind of culture on each of these workplaces that will mske it either more left or right up or down, but they aren't controlled by politicians, just like a hospital isn't told how to treat patients by politicians.
I'd rank BBC News above all US news outlets (and I'll just add I'm neither from the US nor UK).
There are some prolific users who… while their comments are of normal quality, trigger a strong desire for me to respond. I’d rather simply not see their comments (and reply trees) at all.
If you want to roll a quick version of your own, once you've identified rows containing comments you want to block, you need to hide all subsequent rows which have a higher indent. I see there's now an "indent" attribute in the DOM which would make this even easier.
[1] https://github.com/insin/comments-owl-for-hacker-news#commen...
I've never understood the hate for w3schools. No, it's not MDN, but it's not offensively bad either, and it's been a helpful reference from time to time. If w3schools comes up first in a search, I trust that it probably has the answer to whatever very simple question I must have asked, unlike something like Pinterest, which will just be spam no matter what.
They provide real examples for real-world uses right on the page.
I still prefer MDN. I trust it to have up to date information. It was always very good and it has also improved, including the interactive examples.
I loved their Try it Online!
It's more a historical thing. Over a decade ago they were always on top, with often wrong/misleading/bad practice/insecure/simplified stuff. And lots of people thought of them as some kinde official entity because of their name similarity with W3C / w3.org
It's much better now, though. But for lots of old school frontenders the reputation stuck.
In your opinion. W3fools only pulled an “it’s ok now bro, don’t sue us” update. Anytime I give it a go, I’m still presented with inaccurate and harmfully incomplete results.
Want to learn from a “for dummies” website? Be my guest. But my hate is fully justified.
w3 usually just gives me the answer i need to get shit done asap.
mdn is a technical reference. useful. but irritating when you just need a quick explanation.
w3schools = :/
mdn = :O
Given the choice I’d probably block w3schools just so I save those couple seconds a month when I click back and in to MDN.
Same. It's been my go to reference for SQL these days. Unless there's some alternative that I've missed out on