Project92 and the Fediverse - A Smarter Battle Plan to Protect the Open Social Web
www.timothychambers.net
With the upcoming #meta #Project92 Fediverse service, there has been a, well, robust discussion of how to avoid threats looming. Those advocating mass-preemptive defederation make three cases for it. ➡️ To avoid data mining … However, defederation does virtually zero to avoid any big tech entity scraping all the Fediverse public social graph today - Want proof? See here: is.gd/q8U2pv But what if they merge that Fediverse data with their own internal data from IG isn’t that worse than just scraped data?

It is expected to be 2-3 months before Threads is ready to federate (see link). There will, inevitably, be five different reactions from instances:

  1. Federate regardless (mostly the toxic instances everyone else blocks)

  2. Federate with extreme caution and good preparation (some instances with the resources and remit from their users)

  3. Defederate (wait and see)

  4. Defederate with the intention of staying defederated

  5. Defederate with all Threads-federated instances too

It’s all good. Instances should do what works best for them and people should make their home with the instances that have the moderation policies they want.

In the interests of instances which choose options 2 or 3, perhaps we could start to build a pre-emptive block list for known bad actors on Threads?

I’m not on it but I think a fair few people are? And there are various commentaries which name some of the obvious offenders.

  • Kichae@kbin.social
    122·
    1 year ago

    I don’t get what would defederating with Facebook-federated instances gives you, though.

    Site A hosts communities that serve vulnerable people. They see Meta as a threat to those vulnerable communities, as they are not well moderated, and have no issues with hate speech and harassment, so they defederate.

    Site B federates with both Site A and Meta. They act as a pass-through for content from Site A to reach Threads.

    Bad actors on Threads see content from vulnerable people on Site A and engage with it. People from Site A cannot see the bad actors on Threads doing this, but people on Site B do, and bad actors there get alerted to an opportunity to be proper shit stains. Now, vulnerable people on Site A get targeted by this induced harassment coming from Site B.

    What does Site A do?

    They defederate from Site B.

    The question is just about whether they wait until the harm has been done or not.

    • Zaktor@lemmy.worldEnglish
      51·
      1 year ago

      Defederation is a one-way block of incoming traffic from the blocked instance. I’m on lemmy.world and can still see Beehaw content posted by Beehaw users even though they’ve defederated from lemmy.world, but if I comment on that content it will only be visible to lemmy.world users. Beehaw has protected its communities from lemmy.world commenters, but its content is still accessible by anyone for any purpose. Instances that federate with both sides don’t change this.

      • Kichae@kbin.social
        1·
        1 year ago

        I’m looking at beehaw communities on both Lemmy.world and Beehaw.org,and they’re totally out of sync with each other. There’s the rare post from a beehaw user that breaks through somehow - possibly boosted from a kbin or Mastodon instance? - but for the most part, you’re getting basically none of the content from those communities.

        Because beehaw isn’t sending you any updates.

        Is it that you’re seeing beehaw users who are posting to communities hosted on 3rd party communities? Because that’s absolutely possible.

        And that’s absolutely the issue with federating with sites that continue to federate with instances you’ve defederated from. You’re blocking direct communication in both directions, but there’s a lot of indirect communication going on.

        Like, this is literally the scenario I described.

        • Zaktor@lemmy.worldEnglish
          1·
          1 year ago

          No, Beehaw users posting in Beehaw communities visible on Lemmy.world. There’s no third party interaction on either of those posts (just the Beehaw OP and Lemmy.world comments). Whether or not Beehaw is doing the convenience of sending updates, their content is accessible through Lemmy.world. It might take some action on a user here to trigger a pull, but it’s entirely possible and you shouldn’t expect defederation to prevent an intrusive instance from continuing to get content if they want it.

          I don’t know for sure there isn’t some pathway through another instance causing this, but in my understanding that’s not how federated communities work. There’s the owner instance that has the true version of the content and distributes it around and then local copies on each other server that feed their updates back to the main instance. You wouldn’t ever take a third party’s version of a community because you couldn’t trust its legitimacy.