Spyke

Time to ditch #duckduckgo

In the last couple of months I have noticed an increasing trend of supplying me search results that are completely unrelated to the current query and tie back to my location or previous searches. I can say this with a high degree of certainty this is without a doubt beyond the 100th instance this has happened.

My browser is configured against tracking and fingerprinting (in fact all my devices are) which would make it fairly difficult to retain any data unless they are profiling me.

View original on lemmy.world
infosec.pub

I've also noticed that if you do a search, click a result, then hit back to go back to the results they manipulate the results in some weird way so they're not the exact same results you got the first time. Infuriating

99
akiloureply
sh.itjust.works

Yes! I thought I was losing my mind. I was like "I swear this was the third result but now it's the 5th? Eh, maybe it was the 5th."

27

Odd. I wonder what is going on here because when I search that exact same query I get accurate results.

I’ve had iffy search results before but never anything that bad.

47

I also don’t get the same results. I also have tracking and fingerprinting disabled everywhere.

16

I also returned totally accurate results using the exact same query. I would really like to know what is going on here. This is a common complaint with some people using DDG, that the results are poor, but I consistently have as good if not better results than using Google.

0
lemmy.world

great. i finally get off google and now duckduckgo is bad. i hate the internet.

38
larsreply
lemmy.sdf.org

So like… all Searx instances are independent, but I can save my preferences in a reproducible query string? Or like… how’s this a long-term, easy, good move for me? (I know ≈nothing).

2
lemmy.one

You can do it that way if you grab that URL from the "cookies" section. Otherwise, it can save your preferences in a cookie.

2
larsreply
lemmy.sdf.org

Thanks!

Do all the Searx instances generate and keep independent indexes of the entire internet, or is it more like Lemmy (where, maybe, each instance is a window into the entirety of the index)?

1
lemmy.world

Kagi was my favorite but their pricing model is a little too high still. If it was $5 for 500 searches I would do that but the unlimited for $10 is just a bit too high for me.

2

I get that not everyone has extra cash to spend. I'm very fortunate to be able to throw $10 at a problem. I don't begrudge you if you need to save cash (I have no kids to raise, for example). But I think $10 is fair if you search a bunch. Maybe I'm overvaluing it; I have heard great things about Kagi, but I'm still fucking around on Google cause it's "good enough," even though it's in great decline. Fuck it, I'm a hypocrite.

6
lemm.ee

It has been doing this shit for a while. Returning results based on just one key word out of many keywords from my query.

Lately they have been even ignoring the restrictive +"keyword" syntax and returning whatever shit just so they won't admit their (or Microsoft's indexer) is shit

30

It's not the fact that there are filler results that bothers me, after all there might not be results for everything we may possibly search.

It's the fact that these are links to services which have a tie to my unique identifiers which I deliberately declined access to in all available ways!

3
lemmy.world

What would we suggest as a realistic alternative? Sears? Startpage? Brave? (I am no expert by the way, just looking for tips)

21
lemmy.world

I just switched to using Kagi recently and I like it a lot so far! Theres a free tier for 100/searches a month you can use to figure out if you like it.

24
Vexzreply
kbin.social

Me too. I was so reluctant to pay for a search engine at first since there are good alternatives out there I don't have to pay for. But I just at least wanted to try the first 100 free searches and was blown away by how great it is. It has some unique features like prioritizing or blocking specific domains, lenses and custom bangs. I payed the $10 the same day for Pro tier and 5 days later (yesterday) I even upgraded to Ultimate tier with ChatGPT-4 (called Kagi Assistant). I really, really enjoy Kagi so far. Most probably it's gonna be my one and only search engine for the next years to come.

11
lemmy.world

Are, are you telling me that I can block Fandom’s wiki results? Because I dream of such a future!

3
lemmy.world

Searx sucks for images in my experience, startpage was bought by an advertising company, brave is weird for a multitude of reasons but I don't know how good their search engine is as I've never tried it

However I do think that finding a searx instance that works for you and then using another engine like ddg for images is the way to go... You could add a shortcut like ".ddg garbage truck" for easily searching with for example ddg on firefox

22
melooonereply
feddit.de

Also a bit confused by this. I thought the point of searxng is to combine your preferred search engines. So couldn't I just configure ddg for the image results?

6

Maybe their point was that you cannot filter the images, like by color, size, license, etc

But anyway, if I want to show somebody a picture of something it works perfectly As for the results, they are pretty comparable between searx, ddg, and googlag.

I would even say searx wins on this one because 1. It's easier to navigate and open images 2. The nojs version of the search looks so much better than any other search engine

6

I've been using metager for some while and really like it. They're also part of an initiative pushing for a free as in freedom search index, so that's nice

4
lemm.ee

Try kagi.com

I you have to pay for it, but t's amazing

18

Thanks for mentioning Kagi! I've been looking for some Google/DDG replacement but every single one I've tried had some sort of dealbreaking drawback for me. Google collects all my data and heavily filters my results, DDG provides dogshit results most of the time I don't even bother using it, Startpage is awfully slow, and I didn't even get to understand what Searx is as looking for it led me to some random ugly pages.

Subbed for 1 year 10 bucks plan. Works super-fast on mobile too. Hell yeah!

4

I get the same shit at work where I can't use a VPN: like 7-8 relevant results then basically just ads for local businesses after that.

18
sh.itjust.works

I have searx, and i'm feeling like a god of search, better than google, better that bingbinggo, and anything else Although I had a lot of problems with it when I was hosting it on a RPI and docker, now i'm not, and it's just so much better

15
feddit.de

What was the problem with Docker? I want to host it myself, the docker option seems to be the easiest one.

5
Gooey0210reply
sh.itjust.works

Docker by itself is not a good thing, bad with security, not entirely open source, buggy networking, not very reproducible

The problem with docker was timeout for requests, not sure whose falt it was, but the reverse proxy container was unresponsive pretty often

Also, yeah, RPI was not really fast too, it had 8gb ram, boot from ssd Just right now I have i9 with 64gb ram, and this is stupid fast, and actually for most of the apps the performance bump is like 50%

About docker, really, try to look into nixos, it has a really steep learning curve, but it will worth it, and you will be able to do magic

-4
andruidreply
lemmy.ml

Containers are really awesome, but take a bit more to troubleshoot sometimes. Docker is not the only method to run them either. I prefer podman actually, but K3s is the next logical step for running services in a more powerful setup.

All true FOSS too

1
Gooey0210reply
sh.itjust.works

Podman is better, but believe my words, try nixos. It's like a docker-compose file, but for a system, this is really something groundbreaking

You specify all the system and services passwords, usernames, all the stuff, your wallpapers, directories, keys, everything

And all basic configurations are already unified, so to enable some service you just need to add a line in your main config like services.nginx.enable = true; and it just works with all the bells and whistles (kind off, you can add much more. Even more than in containers)

The services are usually not sandboxes, but you can sandbox them, can even run the same containers

Sorry if you're really not into it, it just nixos feels like a whole new lvl after podman

edit: even like that, I manage all my machines as a fleet with nixos, all from one configuration So I can basically press a button and change all the usernames on all machines and everything will continue working

Instead of adding each machine separately to a vpn, I just press a button and it deploys all the machines with wireguard and connects them all

2

No nix is super cool! I really like the idea that guix and nix in having that system as code from build to deployment. I am not sure yet on how I feel about it for fleet/cluster deployments, k8s schedulers, network patterns like service meshes, ETCD, and operating on labels and cluster state are all super powerful.

I have looked too into using nix to make OCI containers and OCI containers to make flatpaks as well. All where they make sense of course.

2

I have still the same good results in ddg, no matter what device or browser. I have noticed that they have some location bias when ddg (and any other search engine i have tried) has non good results globaly but good resulta locally.

Still better then anything i have tested. Just because there are minor anomalies i will stick with them.

15
Liomereply
pawb.social

It does though? Right under the search bar there should be a switch bar to turn on localized search, and a dropdown to pick different location.
Or is it itself location locked?

16

Nope not locked. I have always on vpn (random countries) anti fingerprinting and DDG happily allows the search localisation to my actual or a different country

3
sh.itjust.works

Yes like many other search providers you can set to all, or whatever, but ddg will always localize your searches.

3

I have the opposite problem.

It is always unlocalized, and even if I set it to Belgium, it still is giving me French and German Sites.

2

On DDG I can type in an exact company or product name and not get a proper result, it's crazy. Google seems to find the stuff no problem, it's really quite frustrating when the search engine isn't even capable of basic queries.

13

Yes, I have the same issue with DDG for something like a year now. I can't use it anymore for basic searches, I feel like the bing api they are using is getting worse and worse.

12
lemmy.ninja

If you’re looking for a new search engine, I’ve been using paid search provider, Kagi.com, for the last few months. They’ve been great.

10
lemmy.world

Slowly but surely I am coming to the realisation that free services sooner or later will drift away from their privacy claims

17

Precisely. Ad and user data supported products will always balance making their users happy, with making their paying customers happy.

10

I feel like ddg has never had good search results

9
lemmy.world

I paid for Kagi and have been super happy with it. If you don't mind paying, I highly recommend it. Not having ads or manipulated results is worth it for me.

9
Sunrocreply
lemmy.world

Oh that sucks, can you turn it off or spoof the location?

1

You can easily spoof the location on most search engines by using a VPN, but I'd much prefer my search engine solely respond to deliberate input from the user

4

I turned JS off other than for specific websites. This includes DDG. It gives me accurate results, although the other commenters don't seem be facing your problem either

8
lemmy.ml

Same here. Every time I ditch Google and go to ddg, I just laugh how irrelevant all the results are and I continue feeding Google with personal data and telemetry and whatnot.

It is true that even Google's results feel much worse than before but it's still significantly better than ddg.

8
Gabureply
lemmy.world

Not for me – if DDG is trash, Google is actually unusable.

11

Me neither. Even Bing is better than google nowadays.

Also, when it comes to piracy, google basically blocks every relevant result.

3

It might be blindly searching for websites that contain any of the words scan, mark 2, pro.

It seems to do that when it can't find a page with all of the words, and the only way I know to fix it is put the whole thing in quotation marks.

6

I feel like Qwant has become the best when it comes to the quality of the search results. My experience with kagi is limited, though.

5
katreply
feddit.nl

That is Google, not DuckDuckGo?

3

I am running OpenWRT and forcing DNS traffic to Mullvad with a fallback to Quad9, besides this happens across different networks.

3
lemmyinglyreply
lemm.ee

Tell me more...

Do you maintain the block list to avoid conspiracies or because these types of websites are riddled with nasties?

4

Mostly to avoid conspiracies. The intended users are people who want to protect vulnerable family members.

Another purpose is to demonstrate that the big social networks could get rid of disinformation if they wanted to - "look what 1 person can achieve, in their spare time", kinda thing.

3

No shit in my experience its also filled with random ad stuff and mainly companies selling you shit

1

Looks like you are using Firefox. Use arkenfox sure, but cut Mozilla off it's 115 server network it uses to track you via FF by using a host deny list, FOSS git clone harden-firefox. You'll have to disable to update ublock origin or remove the extensions line, but it's better to just cut the adverts and tracking by removing it from the networks than by browser interception (slower, loss of performance, still hits your computer). Links included to do that in that repo.

Alternative browers are Librewolf and Qutebrowser. When you really don't want to be tracked for some things use Lynx.

A great search engine replacement is Grasp. It's being funded by Paul Graham, the founder of Y Combinator and although you only get 100 free searches a months, it can come in very handy. The search results it gives you, unlike Kagi which is just a reformat of DuckDuckgo yet with AI, it's results are completely different than any other engine and imo, on point, surely for anything technical.

My general search engine is an envs.net free hosting of Searx. envs.net is a free Linux shell community with many services like blogs, email, matrix hosting, etc etc. If you do end up using their German Searx as main search donate to them, I did.

3
lemmy.world

I tried Kagi and canceled after a week. It's a reformat of DuckDuckGo, a better format for sure, and lack of sponsored links, yet it adds AI too. In the end, it's the same old curated unhelpful results that leave millions of high value boutique and indie sources of information out. Also, it's Orion browser is bad.

Basically ask yourself that knowing all the good writers, content creators went to Substack, yet hardly any search engine gives results from there, why?

1

There isn't going to be a search engine without some type of ai or content agitation tool. Key word search is not enough to make one of these work. Search engines need to sort through millions of web pages, and try to give you the best match for what your looking for via smart algorithms. With Google these algorithms are designed to sell you products and get the most clicks out of you. Kagis profit incentive is to curate good links for your search results. Indie results will always be low on the ranks thats why they are indie. As they get more popular so so their search results. You don't want your search engine feeding you only new and up and coming shit. You want the most relevant search results. Sometimes it's going to take some digging to find what your looking for. UNLESS you want to give up mountains of your data to and hope that the company uses it to serve your interest instead of feed you sponsored bullshit.

1
lemmy.world

I've never seen the point of this search engine or any commercial alternative to google. It's all just varying layers of proxy to Google. You might as well just find a searx instance and use that because it's all the same crap at the end of the day.

1