Ask HN: How to do literal web searches after Google destroyed the “ ” feature?
I used this quite frequently but since Google """"improved"""" it last year (there was a popular HN post complaining about this) it doesn't work anymore. Search for a domain name with quotation marks for example just recombines the contents of the domain and returns a bunch of unrelated content completely cluttering what I am looking for. Until last year it used to return no search results if there weren't any exact matches, which is the whole point.
Does someone have a work around for this phenomenal Google decision?
218 comments
[ 0.24 ms ] story [ 365 ms ] threadsite:example.com and -site:example.com still work, I think.
Verbatim really highlights how almost useless the default mode has become.
There's nothing nefarious here. The integration of &tbs= is deeply tied into GWS; it's written into the C++ substrate, while most of the rest of the server has been rewritten in Java + a bunch of custom languages. Most likely they simply can't remove it - it's permanent technical debt that nobody has the expertise to touch anymore.
Changing the UI is all done in a plugin these days, and is a relatively trivial change.
Thank you
https://serpapi.com/playground?q=The+most+popular+fruit+in+t...
Other useful parameters, you can use nfpr to force it to not correct your search terms
https://serpapi.com/blog/filtering-google-search-and-google-...
I wonder if there is now a gap in the market for some kind of "literal" search engine that makes no attempt to infer meaning on your search terms and simply gives you the closest results? In other words Google ca 2012.
I have used double quotes to limit to a particular _phrase_ as recently as last week. I'm not privy to the improvements you mention, do you have a link?
Incorrect:
site: domain.tld
Correct:
site:domain.tld
The - modifier to exclude a word does not have a trailing space.
Incorrect:
- site:domain.tld - query
Correct:
-site:domain.tld -query
I’m less certain about excluding a verbatim phrase, but I believe you can do as follows.
-“exclude this query string”
While this space rule still applies, so don’t insert a space, although this needs more testing by me to know for sure that this exclusion works properly.
- “this is not correct syntax imo”
https://ahrefs.com/blog/google-advanced-search-operators/
I use Google as a fallback but these days it happens perhaps once every couple months, and mostly I don't get anything out of it.
Google remained better for programming questions for significantly longer (I speculate this may be because Google's own programmers used it, and complained when the results sucked :-)), but now it's not. Not really.
Like you, I still use Google as a long shot, but that's become quite rare.
I sometimes use Bing for Microsoft-specific questions, if DDG doesn't give me what I want. I have the sense that Bing covers Microsoft a little better than the others do. I have no real solid evidence for this, but it seems plausible on the surface.
Kagi for me is a no-no. I mostly browse on temporary containers, and also have Firefox delete cookies on close. A search engine that requires me to be signed in would be a massive pain to use.
Maybe its different in the US or just not optimised in Germany, but the quality of search results isn't comparable.
Query: "All you can eat Kiel"
Google:
1. a map with the top three restaurants for this context in the foreground
2: a list of restaurants, sortable by price/date/hours
3 to end of page: links to homepages of the top restaurants
DDG:
1. a completely random assortment of irrelevant links on the left
2. a semantic web box with a 3/5 rated irrelevant restaurant on the right
3. wild youtube-videos vaguely related to the query between text results
4. images
5. more wild links
that kind of thing. And Kiel isn't a small city, we even had DDG print ads here.
When I try your query I only get a bunch of local German stuff (including a map with restaurants).
I admit, I don't even know what Kiel is so I can't validate the results but despite them being in Germany (and my search having had the country selector set to UK), it looks pretty good.
If I search something more generic like "all you can eat sushi" I get a neat map with a few restaurants that do that, plus a bunch of TripAdvisor and other directory and blog posts exactly about that, all local.
If I try your query again, setting the country to Germany, I get https://yummy-kiel.de/ as the first result, a map with top-three restaurants as the second, and again, a bunch of directories and blogs with all you can eat Kiel.
Seems pretty good to me.
Having a minus act like a plus has been particularly tough (for years now, I believe).
I think it's Tay v2.0 injecting itself into the conversation.
I swear that it used to work for certain strings I'm trying to find now which I was able to find information on and now it isn't even returning, with "verbatim" set, something that is in a very well-known program's documentation. Bing finds like three results. Google has dropped the ball so hard it's embarrassing
If they break a key feature to MAKE a verbatim search happen with literally two keys pressed, and are apparently not even indexing what they used to anymore, they're dropping the ball. Most people know the quotation trick now and are probably assuming it still works
I tried using verbatim searches to make Google decent for me again, but Google defines "verbatim" rather differently than I do.
1 - By interpreting your search, it leads to better "search quality" by having one model say "i think this is what they want" and another execute the search. P90 accuracy is increased at the cost of P99 accuracy.
2 - If you search for a literal string you know exists, you expect to find it. By interpreting, fewer search inputs with literal strings make it to the search function.
3 - Since Google is interpreting more searches, this gives ad-placement a route in to favorably interpret "they want to buy something" even when this isn't the case. This makes Google money.
4 - People that used to use literal searches either stop, learn how, or switch search engines.
5 - After a couple years, business metrics show that literal searches represent 0.1% of queries and make less money, is it really worth investing in? When it was a P99 issue it fell off the radar and now the P999 is lost.
So a series or rational decisions by rational actors leads to a decline of a used feature because of business incentives and chasing P90s at the cost of P99s.
Fantastic
I couldn't get my fiancé's mother to stop re-installing bonzi buddy or whatever that toolbar was called because she 'loves the purple monkey, he's so cute!'
I actually use that, but it has its faults. You get more spam results and iffy sites (e.g. Wikipedia clones). It's also missing some of Google's convenient features (like doing unit conversions and arithmetic).
IIRC, Verbatim mode is closer to the raw results of Google's underlying search engine, before some of the massaging they do. Some of that massaging is bad, but some of it's all right.
Second, I just looked up when this feature was introduced (assuming it was fairly new), and it was in... November 2011. It's been there for the past twelve years. See:
https://www.wired.com/2011/11/google-verbatim-search-back/
https://searchengineland.com/responding-to-complaints-google...
https://www.searchenginewatch.com/2011/11/18/google-introduc...
Thanks for letting us know! It's been right under our noses the whole time -- and it's not like the Tools menu is even particularly hidden, at least on desktop.
I think I also mentioned it a couple of times when the same complaint came up.
The problem I think is that "verbatim" is not a word that one thinks of, so nobody searches for that. Plus it's hidden in a generic "Tools" menu. Sometime you get a link to search for the exact phrase at the bottom of results, but that too is subtle.
Every time I tried it, it didn't work.
It certainly changed a lot on those years, but the reason nobody acknowledges it is probably because it's a coin-toss if Google wants "verbatim" to mean verbatim today.
Quotes actually stopped working (they became a hint, instead of filtering the results) a long time ago, and many people insisted for years that the verbatim search worked. Probably because those tried it on the days when Google decided to use a standard dictionary. Nowadays even those people gave-up.
I just tried a whole bunch of queries on seemingly generic-sounding sentences from old pages on niche blogs and it found every single one.
So I'd love to understand when it doesn't work and if there's some pattern -- like if certain punctuation trips it up, or if it's simply pages not indexed in the first place.
But then, I've seen it work too. And stop working again.
This is just about verifying the behavior, that it's not finding a verbatim result that it should. And it has nothing to do with personalization -- it's easy to run these logged out.
Also, "A verbatim search was very similar to putting every word in quotes" is not an example of it not working. The question isn't whether they're similar -- your claim was that it "didn't work", which I take to mean it was returning incorrect results, that correct results were missing.
https://news.ycombinator.com/item?id=36618104
semantic search + realtime needs = google;
raw searches is ecosia(bing)
and most needs are covered via GPT4 actually. Error debugging + code snippets, general interest explanations, arguing things, evaluating things interactively. Basically everything I don't need hard facts of the last 2 years for.
I think that matters very little. Kagi had excellent quality results for me from the start, in a huge variety of topics in several languages. Their user base is probably 90% American hacker, and I'm getting good results on queries they would never use.
The Google search algorithm from 5 years ago was amazing, why they decided to change it for the worse is something I will never understand! And no I do not blame SEO entirely since that existed 5 years ago, what I am often looking for but can no longer find is information that has nothing to do with any products. It's not ads that I need to page through, but unrelated and bad results that are limited. I do not want to see the same results from page 1 on page 3.
I signed up for HN before it was popular to include middle initials or other fancier forms for the username, I'll update my profile though.
I challenge you to find a group of rich folks who will fund such a thing - and will not simply demand the same profits as current structures do.
I just tried this with ~30 wealthy individuals/family funds that I have great relationships with and literally everyone said some form of the following:
“I am not going to liquidate my current investments that are on IPO trajectory for something that by definition will not IPO, and my investments are to ensure my kids go to private school/colllege fund/etc…”
You need to convince people, who currently think that the goal of making/having money is to insulate them and their families from reality, to instead choose to make/have slightly less money so that somebody else can have an easier/better experience in a way that is still concomitant with commercial transactions.
Why not 'challenge' yourself if you've already done quite a bit of groundwork?
Or at least the comment suggests you've already put more thought into this than me.
I need 200k to actually start what I want and currently have 5k in my bank account and a lot of mandatory overhead (kids etc).
If you’re legit interested I’ll tell you all the details 1:1
Email me from the one in my profile
When you make your account, you're given the option to customize. When you do, you can pick things like color theme and how URLs are displayed. On the right hand side of the page there is a preview of what your Kagi searches will look like.
In my example, the demo Kagi search is Magic The Gathering. I play a lot of Magic The Gathering. I spend most of my time online searching for things related to MtG or brewing decks, second only to things related to software development.
I imagine it's coincidence. MtG is a pretty nerdy hobby and Kagi seems like a pretty nerdy product. However, it made me uncomfortable enough to ask:
Is that what it shows for everyone? Or is there some tracking going on already that is being demoed? It's almost certainly the former given the positioning of Kagi in the search market, but I'd like to be sure.
Commander is a 4-player casual format that's has as much in common with more typical board games as it does with traditional MtG.
Arena is likely what you saw someone playing on their phone. Funnily enough, it doesn't support Commander!
I'd been on the fence but after reading through these threads and seeing a real reply from a founder (its been a while since I've seen an honest, non-PR Speak answer) I am excited to try y'all out.
This area was ripe for disruption with how terrible searches have become, especially on mobile. I'd be happy to see you eat G's lunch here.
They try to highlight small, personal websites instead of the big mainstream sites.
(This was a HN submission 2 weeks ago)
Fuck google, I have work to do. Thanks for the tip! Nice realizing that they've basically been wasting my time for a while now and that there's a decent alternative available.
For recipes and stuff it was fine.
But could you not just pay for a month and try it? I don’t think their plans bind you for a longer than a month.
Kagi is incredible and worth every penny simply for being able to remove the SEO scam and tire fire that is Pinterest from all image search results.
I hope Kagi introduces an anonymous access feature. For instance, it could incorporate zero-knowledge proofs (ZKPs). These are cryptographic techniques where one party (the prover) can confirm to another (the verifier) that a claim is accurate without disclosing any additional information. This is especially beneficial for authentication scenarios where it's essential to avoid sharing extra details.
To implement zero-knowledge authentication for quota API access:
1. Token Creation:
- Each month, users receive a token tied to their identity and quota.
- The token can be split for use on multiple devices using cryptographic methods.
2. API Access:
- Clients present a zero-knowledge proof (ZKP) to confirm they have a valid token and haven't used up their quota. The server verifies this without seeing the exact details.
3. Client Synchronization:
- Each client tracks its quota usage.
- Synchronization can be peer-to-peer or through a centralized, encrypted server to prevent double spending of the quota.
4. Quota Renewal:
- Monthly, old tokens expire, and new tokens are issued.
Challenges:
- ZKPs can be resource-intensive.
- Token security is crucial; there should be a way to handle lost or compromised tokens.
- The system should prevent quota "double-spending" across devices.
- If a centralized server is used for synchronization, it should operate with encrypted data.
This way Kagi would only know who their customers are but not what kind of searches they make.
There's only one solution, and that is that you need to put a bit of trust in Kagi. Compared to the major one, Google, you can chose between one that promises to not store data, and one that promises it does (and does a lot).
It's always a bit sad that here on HN, when companies try to do better than bigger players, there's always people who think it isn't enough. It has to be absolutely impossibly perfect.
You can use search engines like Google without being logged in. When combined with tools like uBlock Origin and Cookie AutoDelete, it becomes more challenging for them to build a singular profile about a user, especially one tied to payment methods such as credit cards.
I genuinely appreciate what Kagi is doing, and I'd absolutely be willing to pay for their service, because if you're not paying for a service, you're the product. I trust companies to uphold their privacy promises, but "Trust is good, but proof is better." ;)
Even once you do that, you have all the other tracking mechanisms that the server could use if it wanted to.
For what it's worth, you can buy a physical Mullvad gift card and use that to create a very anonymous account for VPN use.
Even if you buy your gift card from a major online retailer, it comes from a stack of gift cards, nothing tracks which one was sent to whom. You can also exchange gifts among friends.
I don't think it's true. I can immediately see at least two ways how it can be done without identifying the user.
1. Each user gets X tokens at the beginning of the month. When searching, user supplies a token, which is immediately burned. The token does not contain the user identity, just signature validating it's a valid token.
2. Variation of the above: each user gets a token good for X searches at the beginning of the month. When searching, the system will return a token good for N-1 search each time token good for N searches is presented. Again, no need to contain user identity anywhere in the system.
Of course, both solutions have their downsides (sync between multiple devices, stealing tokens, losing tokens, etc.) but it id definitely possible. And I am sure if somebody spent a little time thinking on it, these ideas can be seriously improved to eliminate the downsides without introducing the need to identify the user.
You have to understand that most of us aren't fighting some battle for "perfect privacy," I just want a search engine that works for me, rather than advertisers, at the level of the search results themselves.
But it's also about digital data autonomy. It's not just about avoiding surveillance over sensitive searches, but having control over our data's destiny. Even mundane data, in aggregate, can sometimes be used in ways we can't predict.
In this case though we've have on one hand a product that definitely does aggregate data about searches, and doesn't do what I need very well; and the other a product that could, but does not currently aggregate data, and does an excellent job serving my needs.
And importantly there is no option of a product, available now, that is verifiably prevented from aggregation. Even a VPN unless I disconnect and get a new random IP between every individual search does not provide that protection. (And then browser fingerprints even.)
So while I understand your overall concern, that ship has sailed for search engines and the internet. We're living in a world full of networked cameras that people voluntarily and happily install, of people broadcasting their lives 24/7. The idea of perfect privacy is getting downright mystical/religious.
In the end I think we have to accept in some way that everything we say, write and do is subject to surveillance, and that the government might want to kill you for any reason that you'd have no chance of preventing.
Since you are interested in cryptography, there is a discussion on Kagi feedback site along the same lines as your idea, about possible ways to achieve this without the need for cryptocurrency. [2]
[1] https://blog.kagi.com/accepting-paypal-bitcoin
[2] https://kagifeedback.org/d/653-completely-anonymous-searches...
Also, I found this link [1] in the thread you mentioned. They seem to have implemented something like that.
[1] https://metager.de/keys/help/anonymous-token
However there is no technical way of proving it. So cryptocurrency and cryptography are ways to achieve anonimity from a perspective of a user, regardless of what we are doing.
[1] https://kagi.com/privacy
This reads and smells like ChatGPT / AI.
There are sooooo many other ways to fingerprint than an account.
Oh look, this MacBook with X by Y resolution from this IP address has had 100 searches for the past 2 hours. Oh no! He switched to incognito.
Signed up for Kagi today and have been looking for a permanent gmail solution.
Had this bundle been available today I'd have jumped on it.
I'm considering signing up; it would be one fewer service I'm relying on google for.
Kagi is also working on removing the 1000 limit on the 10$ subscription and offering unlimited searches.
Switching from the 5 $ plan to 10 $ was super smooth by the way, so if you want to try for less $, the 5 $ for a month is enough so you can get used to the product and know if you like it or not (and that's besides the 100 free you get while signing up for a trial).
I usually only need to understand a concept, not understand if the personnel and company names it made up actually exist
everything else I use google for are just addresses
so I’m wondering if a paid search engine would shift my behavior back to search engines, or if that ship has just sailed
For example:
nyt crossword
cheap iem reddit
starbucks near me
M7FFALP
Likely, a good search product in the future will be a combnation of both.
> LLMs can not yet replace good web search ... think navigational queries, shopping /reviews, location aware
amusing.
You should know by now that LLMs will and do lie in subtle ways that are not apparent to non-experts. Using them to understand complicated concepts is a great way to "learn" incorrect information. To be fair, the same can be said for humans, but humans are worse at bullshitting.
I really just need to converse about a topic for more inquisitive-ness and to form structured thoughts
It will tell me if I’m conflating concepts, before bullshitting about the ways theyre different. Thats fine, my blind spot would have been that I was conflating a concept for the next decade.
In that regard its the same or better than a human
I don't need it to be the source of truth, I need it to be conversational. It can make urban legends just like a person does, I don’t care, just give me a way to talk about a concept and decide if I want to learn more and it does that extremely well
I use search a lot in my workflow and my usage is showing 2k/month. I expected it would have been 3-4k.
Fingers crossed, but I have a good feeling about it. If it goes well the pricing seems fair.
The ability to essentially "weight" particular domains (pin, block, or anywhere in between) has saved me so much time. There are certain searches I do (music-related in particular) where I always want particular sites (metal-archives, bandcamp, etc) to be the first results, and having that as an option is great. It means that searches that I perform often have a result within the first 1-5 results that is exactly what I want.
No ads, way less SEO spam, and the ability to completely remove domains from results if I think I need to tweak it further. For most of my searches I previously used Google for- Kagi makes Google's results look laughably bad.
I've also been using it at work for tech searches (linux, redhat, etc) and it has saved me time there too.
I use a "family" account- and have one work account and one home account that way I can have different settings for different environments (would be neat if this could be built into non-family accounts though... like "personas" or "profiles" or something...) because I'm overpaying a bit to have the two account setup and don't reach the search cap. I think I'm okay with that though, because having the cap so high means I've removed the "running out of searches" anxiety from my usage of the service.
Search engines should be able to support even those who are not SEO experts and not the first ones to arrive and sit on specific keywords. What I mean under that, if you have multiple good, exhaustive answers for a query, why not offer varying/random good results, so every link would have a chance? Let people break out of their bubbles.
Same power pyramid scheme. Yukk.
Google's done a lot of "improvements". I hate to say it but its quote feature has been broken for a decade. You're only now noticing?
I've tried using other search engines. I've settled on DuckDuckGo. It also does not have a working literal-quote feature. But it's much less infested with SEO garbage.
This post reeks of Kagi spam to me.
Sorry guys, your product failed, no one is going to pay for search no matter how much Google sucks. Move on.
After seeing the post and trying Yandex it was absolutely right, it’s what Google search used to be.
Now whenever I use Google and it’s just a list of ecommerce adds or content farms duplicating the same content without substance, I head to Yandex and get the type of results I used to get from Google.
That's not my experience? The quotes still seem to work for me? Do you have a specific example? And / or can you point to said HN post?
I spend some time reading a random website until I realize the sentences i was looking for is not there. Useless and frustrating.