On doit apprendre le français (ou l'allemand) pour rechercher le web sans indices Américains? Défi accepté!

FWIW, I gave YaCy a try a while back, and I agree with the article on that one. Shit tier results that make ancient AltaVista look good. Might be fine for intranet search. I like the idea of its distributed hosting, but pass on this one.

Other poster mentioned SearXNG, and while I haven't delved into that too much, it's probably worth a check. Pass on YaCy.

Consider self-hosting SearXNG, which can aggregate results and filter.

Same. I use Kagi because search is an essential function of my job and I can’t extract decent results from Google anymore, but if there were another engine with equally good results and a better ethical track record I’d switch.

(There isn’t. I’ve tried Qwant, Ecosia, DuckDuckGo and a handful of others. Was not impressed.)

indexes is valid: https://www.merriam-webster.com/dictionary/index

The article addresses that in a footnote

This is one part of what I wanted to convey, but for example - bing has some results - they uprank sites they are paid by - then send to ddg - ddg again adds ads - sends to you. /you are effectively using bing with better privacy (I don't mind ads, I can recognise/avoid them). And other problem with ddg is the thing that I brought up - bing is planning/already underway (depending on source of information) to scratch its search api, and mostly replace with a "ai powered api". Most users of bing api would have to find alternatives.

Same with searx, you are more private (assuming you use a public instance), but it is not great for open web. the up/downranking done at source by bing/google can not be avoided (it can be done, afaik there are some scripts/extensions to hide results from certain sites, but you still can not uprank (can not uprank something that was not even sent to you)), and it hurts smaller websites, like articles or forums that are not big platforms or doing heavy seo.

One should most definitely use more than 1 search engine. You can even find interest specific engines.

In this particular case, have you tried checking out uncensored models, maybe trained primarily for west locally, maybe some mistral model. I am not saying western models would give you correct or better or unbiased answers, but if you use multiple sources, you can probably form some results yourself. you can possibly keep a combo of smaller quant of small models (8b should run reasonably if you have 16 gib memory, even 12-14 B models can work, and you can download 2-3 models from different providers)

Side questions: Are you allowed to use huggingface in china? similarly, is all your lemmy life through vpns, or is lemmy small enough not large enough to yet be censored. Also, if you do not mind answering, how is the infamous firewall implemented? is it a blacklist banning certain stuff, or whitelist allowing only selected stuff, or some combinations (like for example, (and I am not accusing china of censoring wiki here, I just do not know enough, and just trying to form a hypothetical example) ban whole wikipedia, but whitelist only certain articles, or whitelist all of wiki, and ban only certain articles, or something like ban english wiki, then whitelist some stuff, but keep all chineese wiki whitelisted and then possibly block some? Also, how is vpn usage treated by government? is it discouraged, or encouraged, or if you use it a lot, and possibly mix some irl activity and vpn activity, is it pursued in some ways?

Sorry for asking so many questions, but I was just curious. In any case, if you feel uncomfortable, do not answer any question. Thank you

Ain't no tree getting planted from a <$0.01 brainwash.

I'm enjoying the results of Kagi quite a lot

I am not gaoing to say anything on ecosia here, since i do not know much about them. All I can say is, most of the ad revenue is going towards bing api pricing, then their servers and other costs of doing a business (for example, rent, electricity, regulations, etc), then their employees, then if they have some amoun of oney remaining, they would possibly use it to plant it. Most people doing business do not immediately want to use their profits, and generally invests parts of profits back to build some runway (ad revenue is not consistent), invest in employees (insurances or incentives), or payback any loans. Even if they have some vc funding or recieving grants, they would have to profitable by themselves, and then only can they plant tress. And keeping this in mind, my guess would be that they are not planting that much

I also did use yacy for some time, but problem is - index is too small. I could within weeks visit each and every page indexed by yacy, and that is not a good thing

i have tried them, but can not use them as first either. but it is fun to sometimes search stuff on it and it goes off the rails

fair point, you currently have to compromise on either "ethics" (a bit stronger word than I would like to use) or result quality. My only try is to promote independent engines in hope to improve them such that market gets more competitive

thanks, fixed

Ddg is bing with a lot of the tracking removed. It’s good for images and quick news searches (remember to add “-msn” otherwise microsquash will fuck that up too).

I can only suggest you to try the engines in the linked article, maybe something serves your wants better

what does that mean