Ask HN: Unbiased Search Engines?

A single shell script that provides command line search using a variety of different search engines; some are www search, some are site-specific. The results are processed with a filter written in C that outputs SQLite3. The SQL database can be used to create simple HTML or whatever I may need. Thus, if I want "meta-search" I can create this easily: HTML pages that combine SERPs from different sources.

Any "search" controlled by a so-called "tech" company has a bias in favour of advertising services.

To get an idea of what is actually on the web, take a look at the .com zone file. An enormous percentage of websites are unreachable via search engine. This does not mean they have no value, it only means they have no value to the sale of advertising services, e.g., they are not popular. Hence search engines are biased against listing them in results.

DDG got crappier when they stopped indexing the Russian search engine Yandex in my opinion. If there's anything I'm having trouble finding on DDG, I just search for it on Yandex now

No such thing as an "unbiased search engine"

And that is not a knock on Bing, DuckDuckGo, Google, Qwant, etc

It is a fact - every search engine is making some kind of decision about `"what is" on the web`: they have their own crawling cycles, internal rankings, ability to reach content, etc

I use Qwant most often on my personal MacBooks

Bing on my iPhone and iPad

DDG on my wife's iPhone and iPad

Google at work (because we are defaulted to Chrome with limited ability to change defaults)

Do they all return "the same" results? More-or-less ... ish

You could run something like the meta search engine searX yourself, and give it the list of other search engines to hit and then return results to you (I did for a little while ... but decided it was not yielding better enough results (for me) to keep going vs just hitting the underlying engines