Following up with my quick glance over web browsers, let's look on second the most used internet-related thing: search engines.
Of course my glance will be specific, so first I will list initial conditions:
(1) No AI shit
(2) Not being a proxy of other search engine
(3) Greatly appreciated if results are reasonable
(4) Appreciated if engine does not spy on me
I am not going to overview absolutely every option as I genuinely might not be aware of some of them.
Why such initial conditions?
First, search engine is meant for finding websites and dassit. No AI prompts are needed. Second, if it is meant for finding websites, then it should pluck out something what makes sense (but it does not have to be something what I was looking for at once, I do not expect miracles). Third, if engine just uses other one results, then I can use the original. If I would need proxy, I would use proxy. Fourth, I understand that people like to eat bread and pay bills, but I do not like being snooped on.
Of course one can skip AI part, use Tor and so on, but this is just my view on this thing. YMMV.
Google - any explanation needed? AI, results are less than mediocre.
Bing - same as Google.
DuckDuckGo - they started pushing AI shit and are Bing proxy.
Brave - has have AI earlier than other guys.
Startpage - Google proxy.
Wiby.me - good choice when you are looking for something more "technical" to look for or you look for old websites. They use their own index.
Qwant - partial proxy of Bing, never used it.
Right Dao - results are bad, though no AI and is not a proxy.
Marginalia - random check says that it is not able to find anything, even a website when provided full address...
Mullvad Leta - recently released to non-Mullvad users, though it is a proxy of Google or Brave.
Kagi - uses AI.
Yacy - sometimes do not respond on SearXNG calls.
Alexandria - not available.
I am aware that there are more options, just wanted to go through big and medium ones. What do we have? Well, we slowly make round and probably it makes most sense to return to webrings, bookmarks and forums. Of course, if you are fine with AI, being spied or you use magic spells to replenish corpo tracking then use what you want. IMO, if you want to use something what is JUST a search engine, options are limited.
My take on self-hosted metasearch enignes - at the end of day they are still proxy to Google, Bing, et cetera. I was running SearXNG locally, but after removing shit, the only engines responding were wiby.me, Yacy and Wikipedia. LibreY, 4get are nice thing as long as you remember that they rely on big, fat fishes (taking 98% of oceans).
Probably what would be the most "moral" solution would be to run own index and scraper, though this needs technical preparations. To Be Investigated.
