Title. I have seen this issue in the past, but I just stumbled across a post where the difference is massive (as of today, 14 January 2024):

316 upvotes and 173 comments on lemmy.sh.itjust.works

https://sh.itjust.works/post/6400367

vs

-13 upvotes and 7 comments on lemmy.zip

https://lemmy.zip/post/3556131

for no apparent reason (i.e. lemmy.zip doesn’t seem to be blocking the instances the missing comments are coming from and those instances aren’t blocking lemmy.zip either). What’s going on here?

Is the issue solved as of today?

P.S. Here’s how the post appears on other instances

  • Ramin Honary@lemmy.ml
    link
    fedilink
    English
    arrow-up
    8
    ·
    edit-2
    11 months ago

    I have no expertise in how ActivityPub works, but my guess is it works something like this:

    You make a post on instance A, you can view it on instance B. People all over the fediverse, instances C through Z are voting on your post.

    The venn diagram of other instances with which instance A and instance B federate overlap but are not exactly the same. There are instances federated with instance A not visible to instance B, there are instances federated with instance B not visible to instance A.

    The vote count for a post is counted by instance A and B based on which instances with which they federate. That means instance A is counting votes from instances that B cannot see, and instance B is counting votes from instances that A cannot see.

    Where the Venn diagrams overlap, both instances will agree on the vote count. But the vote counts coming from instances that they don’t both federate with will cause a difference in vote count.

    • deweydecibel@lemmy.world
      link
      fedilink
      English
      arrow-up
      2
      arrow-down
      1
      ·
      edit-2
      11 months ago

      Really feels like the disjointed and manipulatable nature of voting across Lemmy is going to become more of a problem as more instance admins learn how to abuse it. There are already popular instances that won’t federate downvotes, and that has the effect of making the scores on their instance appear higher overall, which in turn makes it look more active, as well as manipulating the front page content in a non-democratic fashion.

      This lack of a shared reality with the scores is going to hamper lemmy as a reddit alternative if there isn’t some kind of standard or attempt to foster a “true score” across all instances.

      • Ramin Honary@lemmy.ml
        link
        fedilink
        English
        arrow-up
        2
        arrow-down
        1
        ·
        edit-2
        11 months ago

        This lack of a shared reality with the scores is going to hamper lemmy as a reddit alternative if there isn’t some kind of standard or attempt to foster a “true score” across all instances.

        No, not really. You should keep in mind that the vote scores on Reddit and other similar forums are fake scores as well. All forms of online voting are fake to some degree.

        Reddit also does various manipulations during vote counting prior to showing you a score. For example, they may not count scores from accounts that are too new or has too little other activity, in order to avoid scores manipulated by sock puppet accounts. And it is entirely possible, maybe even likely, that Elon Musk directly controls the vote count on X/Twitter posts to his liking.

        So I don’t think voting is really all that important anyway. It may be important on centralized websites like Reddit, but it has always been something of a lie. It is just that on a federated forum the lie becomes more obvious.

        Personally, I never take voting on the fediverse seriously, and I don’t think anyone else should either. I mean, I do upvote, but I use it more as an acknowledgement, more than a means to try to boost someone’s comment over others.