Their design is making it easier to see what the mods/admins think is wrongthink then remove from there. Or if you don’t interact how they want. Then there’s the poorly written code.
Late here, but in this particular case someone using Lemmy spotted a new Piefed account with a username they don’t trust and banned them. It had nothing to do with Piefed admin tools or Piefed itself. The communities the user was banned from are based on Lemmy.world.
Since Lemmy existed before Piefed, pretty much all the bot accounts were here. So their current system uses a database of flagged Lemmy accounts and if the name matches someone else who joins, then it automatically gets flagged too. I think they also check IP addresses at the moment, though I don’t know if they gave up on that or are still experimenting with it.
Oh, for clarification - they’re not automating assessment of user interaction, rather of users who join, and afterwards checking that username to a database of flagged users on Lemmy. It’s to speed up the sign on process a bit more.
Since it’s possible someone else might just happen to pick the same username as that in the database, then it’s possible to appeal and mention you’re not that person on Lemmy.
They’re automating assessments of user interactions on Lemmy as an anti-bot measure?
Their design is making it easier to see what the mods/admins think is wrongthink then remove from there. Or if you don’t interact how they want. Then there’s the poorly written code.
Late here, but in this particular case someone using Lemmy spotted a new Piefed account with a username they don’t trust and banned them. It had nothing to do with Piefed admin tools or Piefed itself. The communities the user was banned from are based on Lemmy.world.
Kinda.
Since Lemmy existed before Piefed, pretty much all the bot accounts were here. So their current system uses a database of flagged Lemmy accounts and if the name matches someone else who joins, then it automatically gets flagged too. I think they also check IP addresses at the moment, though I don’t know if they gave up on that or are still experimenting with it.
That just sound like a terrible idea and as they said for the user.
The automated part was the bigger issue, but it sounds like there’s probably not much assessment to have automated then.
Oh, for clarification - they’re not automating assessment of user interaction, rather of users who join, and afterwards checking that username to a database of flagged users on Lemmy. It’s to speed up the sign on process a bit more.
Since it’s possible someone else might just happen to pick the same username as that in the database, then it’s possible to appeal and mention you’re not that person on Lemmy.
But they are? That’s part of their mod tools to show those they want to allow.