Through the years I tried many different search engine. And unfortunately except google no one has ever been as accurate. I’m trying again duckduckgo and I have not been disappointed so far.
@tkanos@twtxt.net Well, I think the same of Google. I didn’t like DDG so much, so I was looking for options.
Currently I’m very happy with Qwant Lite, Ecosia, and Marginalia. I was using Kagi as a beta user (I think it’s now 10 USD/month) but I went back to the ‘free’ options.
I didn’t DDG either, but that time it’s being good.
- I don;t like Qwant (and behind the scene it uses bing)
- Ecosia long time ago was bad, maybe it’s becoming better (i will re have a look a it)
- Never heard about Marginalia or Kagi ??? What are those ?
@tkanos@twtxt.net I used DDG for a while. Switched away from them because they tracked clicks. I’m not sure if they still do.
Then, I used Startpage which uses Google results. I switched away from them because they kept thinking I was a robot and making me solve a captcha.
Now, I’m on Brave Search. I don’t like their browser much, but I think their search engine is nice. They have their own crawler, which isn’t common. The results are usually pretty good, but when I’m trying to do a per-site search I switch to a Whoogle instance.
Marginalia Search is a search engine with their own crawler that prioritizes simple, readable websites.
Kagi is a paid search engine that, apparently, doesn’t spy on you. However, all your Web searches are tied to your real identity because you can’t pay anonymously.
I used to like brave search, but at some point it was not accurate enough (for me) I left.
@tkanos@twtxt.net Can you define “accurate” in terms of web searches and what you expect? 🤔 As some may know I hacked together spyda.dev one weekend, but its a rather shit search engine really 🤣 I’m always curious to understand what would make a good search engine without spending Billions of $$$ and selling your soul 😅
For instance I kind of like what @mckinley@twtxt.net linked to:
Marginalia Search is a search engine with their own crawler that prioritizes simple, readable websites.
I kind of want Spyda to be this… And recent activities and discussions also made me realise there’s no reason Spyda couldn’t use chromedp as part of it’s “scraping” as long as the content that comes back is in some way or another “reasonable text” that can be indexed.
Accurate was most, when I was unable to find what I wanted in the two first page, so I went to google did the same research and there it was. So after a while doing double search I gave up in brave search. Even if I loved to read their paper.
@tkanos@twtxt.net Hmmm I see 🤔