• Milx@lemmy.world
    link
    fedilink
    arrow-up
    2
    ·
    1 year ago

    Is “how many people from your own instance subscribe” even a useful metric? I don’t see what value it brings, I don’t care how many people on there happen to also come from my instance, I just want to know which instance has the most active version of this community.

      • Milx@lemmy.world
        link
        fedilink
        arrow-up
        2
        ·
        1 year ago

        What do you mean by “established” though? Is a community with 80 people from my instance and 500 more across all other instances less established than one with 250 people from my instance and 30 more across all other instances? If so, how? Legitimate question - I’m new here and it’s possible there’s a good reason to care, but I can’t see one.

    • AbouBenAdhem@lemmy.world
      link
      fedilink
      arrow-up
      2
      ·
      edit-2
      1 year ago

      That would require each subscribing instance to report its number of local subscribers to the instance hosting the community, which would then add up the numbers and report the total back to the subscribers. That seems like it would be easy to manipulate, where one malicious instance could misreport local subscriber numbers to change the apparent relative popularities of different communities throughout the fediverse.

      • Empyreus@lemmy.world
        link
        fedilink
        arrow-up
        1
        ·
        1 year ago

        I mean if that’s what we worry about then essentially everything could have that problem. What’s to stop an instance from artificially increasing their upvote counts so their posts are always at the top.

        • AbouBenAdhem@lemmy.world
          link
          fedilink
          arrow-up
          1
          ·
          edit-2
          1 year ago

          Sure, but the amount of work involved scales differently. To inflate post or comment rankings by a thousand votes, you’d need to create a thousand separate fake upvotes per item, whereas for subscription totals you misreport one number and the effect is permanent. That would make malicious actors harder to detect.

    • TeaHands@lemmy.world
      link
      fedilink
      arrow-up
      1
      ·
      1 year ago

      It’s such an obviously weird stat to use, I’ve been assuming there must be some technical reason why it’s more complicated than we might think.

      • madjo@geddit.social
        link
        fedilink
        arrow-up
        1
        ·
        edit-2
        1 year ago

        I’ll give you two links to the same post but on two different Lemmy instances:

        https://geddit.social/post/17196

        and

        https://beehaw.org/post/659342

        It’s the exact same post, with the same user who created it and everything.

        But on the Geddit instance, you have several comments answering the question. On the Beehaw instance, you only see my test post (which you won’t find on the geddit’s version).

        And truth be told, I have no idea why there’s a difference there. Both servers are federated with eachother:
        https://beehaw.org/instances and https://geddit.social/instances
        And yet, there’s a difference.

        edit
        One thing I could think off is that maybe it’s not federating because there are no subscribers there. I’ll try that next.