I thought paid Kagi would be a real search engine. But it's not. And Kagi's browser is closed source so that's a no go too.
I thought paid Kagi would be a real search engine. But it's not. And Kagi's browser is closed source so that's a no go too.
I use forums, wikis, and content aggregators (e.g. Reddit) to discover related topics. If a search engine returns too many results, I refine my search terms.
1) Before Google.
2) Google before ‘08.
3) Google after ‘08.
1 and 2 were both pretty good for just exploring.
It does completely kill the anonymity though, I agree. It's strange that in this case they're worse than something like duck duck go. Because they don't require payment they simply have less data on you (especially confirmed data). I'm sure kagi protects it but the data you don't even have is even better protected.
I don't need to trust anyone not to share information they don't even have.
They should really allow anonymous payments like mullvad does. There you're just a random number and you can pay with crypto without any addresses.
It’s not “browse search results” but more “curated stumbleupon”.
You can also change your search lens from just generic “web” to “small web”, “forums”, “academic” etc or create your own lenses.
I don’t think these answer your particular browsing pattern, but I for one am happy that it doesn’t return hundreds of results. I feel like that is kind of the point, even. I’d rather get fewer, but better, results and have it just say “look buddy, there isn’t anything else”. Plus, I’m not stopped from just adding `!g` to the query and getting 1000 garbage Google results if I want.
Provide a random one if this matters to you? It's not like they have any way to check.