• 3 Posts
  • 29 Comments
Joined 8 days ago
cake
Cake day: November 11th, 2025

help-circle






  • Thanks for making Blorp, it’s my favourite Lemmy app I’ve tried! And the lack of post flairs is something I’ve noticed, so I might try PieFed out.

    If you want feedback on some more (very minor) things I would love to be fixed, here’s a list of things I’ve noticed:

    I posted on the Blorp community yesterday about wanting a “read all” button for notifications (mostly because this post got a lot) so that would be really nice.

    Other than that, it’s mostly just tiny things. Apparently Lemmy has a backend way to see user karma, so maybe seeing that would be nice to add as an option?

    Maybe a button to copy and paste a post’s body text on mobile like on Reddit, or even a “save image/video” button? This is really minor.

    And maybe an easier way to see the upvote/downvote ratio? (like 87% upvoted as an example)

    Maybe this isn’t possible for some reason, but would a way to create a community inside of Blorp be possible? I don’t know if Lemmy works like that.

    Thanks again for making this app, especially the fact that it’s on iOS. I hope it’s not too much of a time/financial burden because it’s great.

    (All of this is from iOS by the way)


  • This is more of a recent thing.

    Since the API changes, everyone’s been pretty much forced to use their first-party app, so they can collect even more data on you. But, even without it, they still collect a comically huge amount of data like how long you hover on a post or how fast you type on your keyboard.

    That combined with the seemingly sudden leap in AI within the last 3-4 years has allowed them to employ AI specifically trained to detect evaders.

    If you want to keep using Reddit, be careful, because second chances are very hard now!











  • I don’t even think the evasion detection is that complicated.

    Barring fingerprinting and IP shit, let’s say they index every single user that spends more than 25 hours on their site per month to lower computational load relative to if we did every single user.

    For each of those users, they might record 10 of their most used “niche subreddits” (those with under 200,000 members, niche because most reddit users are in the non-niche subs so it’s not conclusive enough)

    From that data alone, if a user happens to be into, say, Golden Age Minecraft, 3D Printing, Stampylongnose, Namesoundalike memes, Drakethetype memes, and be in a Saskatchewan local sub, ON TOP of the account being relatively new, this is enough evidence for Reddit to be 99.9% confident that you’re the same user as someone else with that data—and if that someone else with that data is banned, you’re gone too!

    That’s how I think they do it. So, you have to completely change your usage habits on top of all device fingerprinting and IP stuff. Near impossible unless you’re not just terminally online (like me) but eternally online.

    Seems unlikely that they’d do this just to catch ban evaders? They probably were already doing it for targeted ads, so might as well repurpose it for this, too!