Hacker News new | past | comments | ask | show | jobs | submit login
Searx instances (searx.space)
79 points by troydavis on Aug 22, 2020 | hide | past | favorite | 27 comments



To add a searx instance to your browser's search engines, visit the instance and follow the instructions on https://searx.xyz/about ("How to set as the default search engine?").

For example, to add the instance searx.xyz to Firefox, visit https://searx.xyz/ , click on the "..." icon to the right of the URL bar, and click "Add Search Engine." You can then choose searx while typing a query or can make it your default search engine (see about:preferences#search).


On my FF I'm getting this:

  Secure Connection Failed

  An error occurred during a connection to searx.xyz. The OCSP response contains out-of-date information.

  Error code: SEC_ERROR_OCSP_OLD_RESPONSE  
  
This instance worked however: https://searx.dojocasts.com/


How do websites get the “Add Search Engine” button? For instance, I noticed I can do it with Gandi but not with Name or Namecheap.


They publish a small “OpenSearch description file” XML document somewhere on the domain, then put the path to that file in an HTML meta tag (which enables browser auto-discovery). Here’s more: https://developer.mozilla.org/en-US/docs/Web/OpenSearch


Very interesting, thanks :)


Doesn’t it violate google search terms? If I setup it locally, would I have to solve captchas all the time afterwards?


I tried to use searx for a while, but gave up due to the lack of decent instance.

Ended up using startpage.com - it might be a tad worst for privacy but a lot better for usability IMHO. If anyone is aware of a better alternative, I'd love to know about it.


I have been using my own instance for about 2 years now. I couldn't be happier. While researching for my master's, I never felt I was missing out anything substantial.

Setting up your own instance is as simple as mirroring the Git repository. Mine is running in the default 8080 port.

Here's more info on how to do it: https://asciimoo.github.io/searx/admin/installation.html


I've setup my own instance on a VPS of searx (along with git, blog, private npm and a few other bits and pieces). It isn't particularly hard if you've setup an website on a Linux box before.


If you set up your own instance and are the only one to use it, what's the point of using searx and not grogle directly?


Because my goal is to move everything to self hosting whenever possible. Also there is nothing stopping anyone else from using it.


Right, the way I see it, searx model is flawed.

I want to host my personal instance. But doing so doesn't make any sense since if I'm alone using it. So I'm better off using Google directly, unless someone can search whatever he wants via a service I'm hosting :/


Well if you use google directly, google can track what you are doing. However if you use your own searx instance that is very difficult. I also host my instance publicly (I've not added to the list because I am using the cheapest VPS from vultr) and for some things I get better results than Google, DuckDuckGo etc.

So as with many things it is swings and roundabouts and is determined on what you want to do. I wouldn't call the model flawed, it just seems that the benefits of it aren't what you want. Which is fine.


> However if you use your own searx instance that is very difficult

The whole point of searx is not to mix your google research with other people google research, hence making it harder to track you?

AFAIK being alone in using a searx instance is useless, as google will still track you via this searx instance.


Are there any good FOSS projects that would let me run my own crawler?


YaCy is the only one I'm aware of. See https://www.yacy.net/demonstration_tutorial_screenshot/


YaCy is a decentralized search engine


Try Apache Nutch.


Looks like Common Crawl switched to this from their homegroan crawler. That's high praise!


isn't that what Searx is? https://asciimoo.github.io/searx/


No, it is a metasearch engine but not a crawler.


I wrote a short quide on installing searX on a Linux system back in Feb. 2020

Non-Docker Local Installation of searX on Linux: https://persagen.com/2020/02/02/searx.html

I use it routinely as my default search interface to the web, defaulting rarely e.g. to Google for image seaches.


I love the idea of searx. And ran a searx instance for a bit. Overall good idea. Slow search execution.


Searx is only as slow as your configuration allows it. You have to check the preferences, under engines, and deselect the offending slower engines. Some will take up to 10s to return any result. If you only choose faster engines, searx will also be faster.


Is there any explanation what is this?


It's a search engine that searches other search engines for you with a focus on privacy (e.g. provides direct links to results instead of Google-style tracking links, etc): https://en.wikipedia.org/wiki/Searx


Searx is a metasearch engine [0].

"A metasearch engine (or search aggregator) is an online Information retrieval tool that uses the data of a web search engine to produce its own results."

[0]: https://en.m.wikipedia.org/wiki/Metasearch_engine




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: