1. Reddit sells its api for high and is about to go for an IPO, its economy bases entirely on the data made by the users/communities. It is the work of the public, get robbed by a small group of individuals. A living example of capitalism.

  2. Fediverse isn’t enough to secure the publicity and usage of public data. What if the host of Lemmy instance also releases the snapshots of all the posts and modlogs, everyday, in the form of bittorrent? Only by doing so, we are safe from the host erasing public knowledge and data brokers.

  • Bezier@suppo.fi
    link
    fedilink
    arrow-up
    6
    ·
    10 months ago

    I don’t quite understand what you’re going for.

    How does this protect us from data brokers?

    Can’t you already pull the data of an entire instance via the api?

    If you don’t trust your instance admins, have you considered setting up your own?

    • wargreymon2023OP
      link
      fedilink
      arrow-up
      1
      arrow-down
      1
      ·
      edit-2
      10 months ago

      Can’t you already pull the data of an entire instance via the api?

      You could ask the host through the api to do that, and that’s the same the problem with Reddit. What we have changed here on Lemmy is more instances to choose from, the public knowledge grows with the instance and I wouldn’t expect the host to let you access the data for free as the value(communities) grows.

      There is real value in the data, way more than ever, bc it is the source for LLM.

      • Bezier@suppo.fi
        link
        fedilink
        arrow-up
        4
        ·
        edit-2
        10 months ago

        So you fear that instance admins would start closing their apis. It would’ve been real helpful if that was on the on the original post.

        The api is accessible and I don’t see that changing because it is, y’know, required for federation. If you want backups, you can start scraping or host a private instance that subscribes to everything you want to save.

        Releasing dumps the way you described would be a massive burden on admins, or even completely infeasible.

        • wargreymon2023OP
          link
          fedilink
          arrow-up
          1
          arrow-down
          1
          ·
          10 months ago

          Releasing dumps the way you described would be a massive burden on admins, or even completely infeasible.

          It adds work but it is actually very easy, the texts and images aren’t that big. The burden is eased by more people engage in seeding(hosting) the data.

          • Bezier@suppo.fi
            link
            fedilink
            arrow-up
            5
            ·
            10 months ago

            Well, let’s just say that I doubt you’ll be able to sell this idea to instance admins.

            However, you can become the change you wish to see yourself. You can start scraping today and create your own dumps that way; the apis are yet to be closed.