I’m a long time Lemmy lurker and occasional Redditor. Since the Reddit influx, I’ve watched the frequency of shitty Reddit-type behavior, e.g., combative comments, trolling, and unnecessary rudeness, just sky rocket.

I’m happy to have more content on Lemmy, but I wish the bad actors and assholes would have stayed on Reddit.

Yes, I realize the irony of posting this on a new community that’s basically a Reddit transplant.

  • Lvxferre@lemmy.ml
    link
    fedilink
    arrow-up
    2
    ·
    1 year ago

    We mostly agree on memes and other “just for fun” material: it’s fine if it’s there (I like it too). The only problem is when it drowns the deeper content into a sea of fluff, as it often happens in social media.

    but maybe if you could only use one of those a day, or ten per month or some such

    What if its value decrease with usage?

    For example. Let’s say that the feature is called “fav”. And that “favs” are taking into account, for sorting purposes. Each poster gets 100 “fav points” a day.

    If the person “favs” a single piece of content, that content is boosted by 100/1 = 100 fav points. If the person favs two, each gets boosted by 100/2 = 50 points. And if the person indiscriminately favs 1000 pieces of content through the day, each is boosted by only 100/1000 = 0.1 fav points, so practically nothing.

    This wouldn’t impose a hard limit on how much you can use the feature per day, contrariwise to your idea, but it still makes you use the feature consciously - because you know that favving one more piece of content will make all the others that you’ve favved through the day count less and less. I feel like this could address the fluff principle in a way that simple votes (or boosts, double upvotes etc.) don’t: not using the feature would backfire (the points go to waste), but using it indiscriminately would also backfire.

    I’ve read the Ennui Engine article. I feel like the author touched a good point, perhaps this is all a result of us taking the internet as “it is not serious / real life, then it doesn’t really matter”. This mentality somewhat worked in the 00s? Not any more though. The proposed solution feels unfeasible though, as it expects people to do the right thing, that’s like herding all cats into the same direction; we might need smarter solutions than that. (Even then, thanks for sharing this text, I think that it bullseyes the problem on the descriptive level.)

    Thanks for the nomination in the mag!