I’ve been using different SearXNG instances for a while now. Typically I’ll whitelist cookies in Firefox for whichever instance I choose to use, and I configure it to use Google, Bing, Mojeek, and Brave.
Inevitably several of them will return errors, and when it gets annoying enough I switch instances.
But I’m getting bored of switching all the time.
What’s your approach to using SearXNG? Am I just doing it wrong?
I think the best thing in your case would be to create your own custom instance. I use it from time to time with Librewolf, but actually, nowadays I prefer to use Brave Search. It may be less private, but it is much simpler, although I have clear that in the future (hopefully near) I will create my own instance with searches in brave, ddg, startpage, qwant and mojeek (I will also put video search directly in my also future instance of invidious and things like that, or for example access to the public answers of the fediverse as it is done with twitter and reddit).
And answering your question, no, I don’t think you’re doing it wrong, changing instances every so often can even be good for your privacy, but I understand that it can be a hassle to adapt SearXNG to you and have to deal with the bugs that sometimes occur.
1 Like
This is why I personally do not use SearX or SearXNG. That, and the fact that I would rather not trust random instance operators with my searches.
We recently had a discussion about removing it, but decided against it for now, for the reasons mentioned here:
That said, I don’t think there’s much to be done with regards to being rate limited, unfortunately.
2 Likes
Thank you for your response.
The problem with hosting my own instance is that I’ll stick out like a sore thumb if I’m the only one using it (and I don’t want to spend the effort trying to get other people to use it).
I might just switch back to Brave or DuckDuckGo.
@matchboxbananasynergy makes some good arguments against using SearXNG.
Regardless, I think its pretty neat, and I’d still like to know how others make use of it.
1 Like
I don’t see it that way, I mean, if you host it yourself on a machine that you can control, you can pick and choose what information does or doesn’t go out of it to the internet.
If you harden your instance and server, I don’t think there would be much of a problem with that, but hey, I can see your point.
I’d recommend Brave, as it has its own search engine, whereas DDG takes its results from Bing. Other interesting search engines could be Startpage or Metager (disclaimer, this search engine is my own recommendation), although I still consider Brave better.
You are right, even if there are things I don’t agree with him, I have to admit that @matchboxbananasynergy is based.
1 Like
if you host it yourself on a machine that you can control, you can pick and choose what information does or doesn’t go out of it to the internet.
Ultimately every link you click on via a personal SearXNG instance is going to be able to build a profile on you, the only user.
Especially in the case of SearXNG where your search queries are forwarded to several (possibly privacy-invasive) search engines. If you have several users on an instance it doesn’t matter so much, since your profile would end up obscured by many other users’ queries.
I’d recommend Brave, as it has its own search engine
I use Brave on my phone since it comes default on the Brave browser, and I think it does the job pretty well.
But… I like the duck…
This is not exactly true, as I mentioned in my reply on the other SearXNG thread which was linked above. The backend search engine providers have no way to determine which link you end up clicking.
1 Like
My point was that websites you find via the search engine will be able to see the URL you’re coming from. If you’re the only person using that instance, it would be all the more transparent.
My point with the backend search providers was that all the queries made on a SearXNG instance (not the links you end up clicking on) are forwarded to said search providers, which they (the search providers) can also use to build a pretty accurate profile if you’re the only user.
I think
This is not the case, your browser should not be sending referrer information to the sites you visit from Searx.
Your second point is correct, but it does not negate any of the benefits I mentioned in my other reply.
Edit: It also occurs to me that you could replicate having many searching users on your instance while keeping it private by simply connecting to a public VPN provider on your SearXNG server, although that does not solve the rate limiting issues.
2 Likes
Just to add another perspective and use case to the mix, I use SearXNG extensively for research. Aggregating results is really useful for literature review because it saves so much time and reduces the chance that I miss relevant research.
You can randomize usage of public instances (“what about trust extension?”) but I’ve recently moved to using a single trusted instance.
Compartmentalization circumvents any worries, for me. Although the information is non-identifying anyway, the activity is already transparently reflected by my publicly available information, so fingerprinting is a non-issue.
Usually, and as far as I know, search providers don’t build any profile based on an IP but on the cookies. But yes, they theoretically can build your profile if you are the only user.