As asked.
I think mojeek is the best in privacy atm but it isn’t really good as a search engine.
you can smash the “submit feedback” button on the page when you get a duff search and send it into us, it’s how we improve: https://blog.mojeek.com/2024/02/major-algorithm-update.html
Will do and i love your ethics.
First time I’ve heard of Mojeek. Why should I trust it more than any other company? Is there anything particular about its economic model or governance that makes it less likely to decide to be unethical?
I recommend startpage.com over duckduckgo. It provides at least the same privacy as DDG while having generally better results because it uses google’s results instead of bing’s.
Edit: The ultimate search engine though, as mentioned by macattack, is SearxNG because it combines privacy with every other search engine and even allows you to mix and match your result sources.
I self-host searXNG, but you can use one of the public instances as well. My understanding is that it is more secure because you’re search results are commingled with whoever else uses the instance, but you also can use something like libredirect to further distribute your search results across various instances for further security
How exactly do you self host your own searxng instance? Also, based on what you said, wouldn’t having your own instance make you stand out more?
Having your own instance can be bad for privacy, as all your searches come from your IP (hosted at home) or the same IP (hosted on a server). They might not be traced to you personally, but you might still get personalized results or your search may still be tracked, depending on how they track you.
That’s circumvented when using it with some or better many other people. But then, you need to trust the admin of that instance.
Self-hosted is easy if you know a bit about servers. You need a domain pointing to a server. If it’s the only thing hosted on that server and you have set up docker on it, you can just follow their instructions here to get it running in less than 5 minutes (assuming you run the default config and don’t customize all of the settings for a while): https://github.com/searxng/searxng-docker?tab=readme-ov-file#how-to-use-it
Ah ok gotcha. So it’s a trade-off of having the instance always up vs privacy? Interesting. Thanks for the detailed info! I keep meaning to get into self hosting lol
No, it’s not „always up“.
There are three main ways how Google, Bing,… can track you:
- When you’re doing a search while being logged in, it’s probably you
- If you’re not logged in, they can set a cookie to recognize you on your next visit (although they may not be able to link this to you, your email address,… but that’s not needed). They may mix your searches with those of the other users of your PC, when those are using the same PC, browser and account (e.g. if you have a family PC with a single windows/Linux account that everyone uses)
- Even if you’re not logged in and don’t accept / delete your cookies, they still see your IP. Depending on your ISP you might have the same Ip for a long time or you might have it rotated regularly. Now they could only track the searches of your household (assuming everyone isn’t logging in and deleting cookies immediately)
With Searxng, they can only do the last variant. But assuming you use a “real” server in the internet (and not one at home), it will likely have the same IP for its lifetime. And if you’re using it alone, that’s the only thing they need to identify you and track your searches. The more other people use your instance, the less useful this kind of tracking gets. Too much noise to identify a single person.
Ah ok that makes more sense, thanks!
Someone already gave an extensive comment about how to set things up so I will skip that part.
Good observation re: self hosting potentially reducing privacy. The way that I keep my privacy during self-hosting is to completely avoid search engines that track my IP address, and then, ideally, although the remaining search engines are less efficient than the likes of Google or Bing, the fact that the results are aggregated hopefully increase the efficiency of the results.
For my default searches, it uses mwmbl, mojeek & qwant
Oh ok gotcha! Thanks, that makes sense!
No one really talks about it and it’s in its infancy, but I’ve really been enjoying Stract.com
I really want to like stract but I can’t get it to give me good results. I typically search for programming related things and it has difficulty pulling up relevant information regarding things like documentation and posts regarding the issue.
That’s fair, I can’t say I’ve daily-driven it yet since I’ve still got my Kagi sub, but I’m thinking of defaulting it on my work computer to put it through its paces. It’s done fairly well with some of my tests, but definitely seems to still be getting refined, I’ve noticed recent information can be difficult to find on it
This is what I use because they take your ad revenue and plant trees. Judge for yourself.
coming up to 20 years of no tracking, 18 of having the first privacy policy to say it: https://web.archive.org/web/20060318104627/http://www.mojeek.com/privacy.html
hopefully a good consensus 😅 at least seen very little of people saying bad
My first choice is actually Kagi these days. I pay for my search provider to have some peace of mind that my search provider isn’t selling me.
I got it not too long ago, and I love it. The results are good, and the features are what i wish all search engines had. Just not looking forward to when they sell to venture capatilists down the line, and everything becomes terrible.
In the meantime, everyone should check out their Small Web intiative. Gives you a random blog or small website. There’s some really good articles I would have never read otherwise.
I used to pay for kagi, but their CEO have some opinions that I absolutely hate about how to handle some issues, and this was a deal breaker for me. Nowadays I’m using searxng
What opinions can I ask? first time I’ve heard this
They announced a few months ago that they would partner with Brave to surface Brave search results too. The CEO of Brave is known to be homophobic. People got mad, and Kagi’s response was that they are too small to be picky, and have to focus on search quality.
And theres the issue someone opened on their page suggesting to include a special card on suicide related searches telling people how to find help, and the CEO dismissed saying it can be just someone curious making the search and he don’t want to “set a precedent” on showing things the user is not actively searching
I… think I agree with that though.
I’m not saying there isn’t a grey area between being socially conscious or full blow libertarian, but I do think that a search engine operates much better when it is unrestricted by societal qualms.
Look how stunted ChatGPT becomes with each new rendition, or how SDXL is far more restrictive in what it can express compared to its earlier SD1.5 models.
I used to use DDG until they started censoring results.
I use Brave Search and Startpage. The latter is basically just a private Google proxy with some cool features and the former is fully independent, private, aesthetically appealing and has nice optional AI features.
DDG really does censor results hard. They even blocked my list of US atrocities from their search results.
Contributing here with another sample point.
I did not find the mentioned link on either search engines’ first page of results.
DuckDuckGo:
StartPage:
I am not sure why a black bar appears when taking a full page screenshot of StartPage on Safari for iOS, but both screenshots are unedited and uploaded as-is.
The screenshot for StartPage is blurred for some reason. Re-uploading again:
E: Re-uploaded a fresh copy after restarting Voyager.
duckduckgo is a US company and the CEO is the previous founder of The Names Database. i don’t know how anyone trusts it when there are so many EU alternatives
So many yet in a thread asking for names you could name one.
Like for example?
Qwant (france-based), startpage (NL-based), searxng (bunch of host options)
honorary mention to Stract which is currently totally useless because the results are so random, but hopefully in a few years it will be a proper alternative. they are indexing the web from scratch in order to not have to rely on existing data from other search engines (which is why the results are terrible right now lol)
Start page was purchased by an ad company a while back
FUCK ME.
Yeah, there’s no winning
gemini search engines are not so bad for likeminded folks with likeminded topics.
ex: <rare linux search> 12 results!
vs searching windows how to pay for windows. 0 results. srry. gemini is small.Gemini as in google AI search? Or gemini as in the protocol?
gemini protocol. thank u for prompting for clarification.
what search engines are there?
gemini://auragem.letz.dev/search/
gemini://kennedy.gemi.dev/
gemini://geminispace.info/
gemini://gemplex.space/
gemini://auragem.letz.dev/search/features/
gemini://tlgs.one/doc/search
gemini://kennedy.gemi.dev/docs/features.gmi
gemini://geminispace.info/documentation/searching
If you use Mullvad VPN, you have access to their Leta search engine
I didn’t know this. Do you like it?
It’s a proxy for Google, so it’s the same search results that you would get on Google.
I use brave search by default.
Not sure why are you being downvoted. Brave search is actually pretty good
Yeah, I believe that the common consensus here on Lemmy is that Brave is ran by an A-hole and shouldn’t be supported. Secondly it’s hated for injecting it’s own ads and tracking into things aswell as the whole crypto thing.
Imagine canceling one of the fathers of the modern WWW because he hurt your feelings…