Comment on Kagi silently removed all references to Google's index from their website

muntedcrocodile@lemm.ee ⁨8⁩ ⁨months⁩ ago

• BaroqueInMind@lemmy.one ⁨8⁩ ⁨months⁩ ago

  It still requires the use of Google/Bing/etc API calls. There’s literally no way to truly self host a web indexing search engine without sacrificing your privacy or paying millions of dollars.

  source

  • hedgehog@ttrpg.network ⁨8⁩ ⁨months⁩ ago

    You can use YaCy, which can be run as an independent self-hosted index (in “Local” mode), where it will index sites visited as part of web crawls that you initiate, or you can run it as part of a decentralized peer-to-peer network of indexes.

    YaCy has its own search UI but you can also set up SearXNG to use it.

    source

    • BaroqueInMind@lemmy.one ⁨8⁩ ⁨months⁩ ago

      I have mentioned this software a while back here in lemmy and someone with actual expertise mentioned running YaCy on noncommercial hardware becomes untenable after a certain duration due to poor quality network, incompatible indexing algorithms, massive database, and search query response.

      source

      • Dark_Arc@social.packetloss.gg ⁨8⁩ ⁨months⁩ ago

        I am not that person, but the only way I see YaCy being useful is as a web crawler for specific sites that you personally find high value in.

        source
• Cheradenine@sh.itjust.works ⁨8⁩ ⁨months⁩ ago

  You can do it, or use one of the instances at searx.space

  Searx is great, it’s all I use, but it’s a meta, there is not a ‘Searx Index’ which is what this is about.

  source

  • hedgehog@ttrpg.network ⁨8⁩ ⁨months⁩ ago

    there is not a ‘Searx Index’ which is what this is about.

    There’s YaCy, which includes a search index (which can be independent or can join a P2P network of indexes), web crawler, and web ui for searching. It can also be added as a SearXNG engine.

    source