Makes sense. Even then a lot of the troll’s comments are upvoted by a few people. There are definitely a few people here who share the troll’s extreme opinions. Thanks for doing your part in removing all these bullshit. I wish lemmy had a way to block account creation by IP addresses. That might stop trolls from making dozens of alt accounts.
We very often see the same username created across many instances, it’s very easy to do and Lemmy has no protections against it. Plus, there are no protections against creating multiple accounts to upvote your own posts (don’t get any ideas 😆). IP blocks wouldn’t work as instances are entirely independent, so there is no sharing of IP info across different instances.
Currently there is at least some level of coordination across instances, though, such as Lemmy.world’s Defense HQ, where instance admins can share info about spammers/trolls so we don’t have to wait for a report from one of our own users. There’s also Fediseer, but this protects against spam instances not spam accounts on mainstream instances.
We very often see the same username created across many instances
Guilty as charged. I’ll say though, there are several legitimate reasons why one might want to do this. I personally use it as a substitute for Reddit’s multireddit feature, by grouping community subscriptions across different instances by theme. As long as users use the same username across instances I don’t think this practice should be automatically regarded as an attempt to sockpuppet. It that was the goal, the accounts would definitely not be using the same username across all the instances.
Personally I have accounts on multiple instances because I wanted to make communities in different languages and some instances focus a lot more on one language than others and also because the SFW instances defederated the NSFW ones. I do not really interact with the same posts though.
I didn’t intend to imply that by using the same username across instances you were breaking some sort of rule. Different instances have different moderation policies, different federation policies, and different intents. Having multiple accounts in good faith should not be an issue and was not what I was trying to imply.
Rather, the intention was to show that we know bad actors do this with nefarious intent. Here’s an example (they show zero comments as they have been banned with content removed - also I think these ones only had posts not comments anyway):
there are no protections against creating multiple accounts to upvote your own posts
They don’t even have to be real accounts. Lemmy uses the ActivityPub protocol, and nothing’s stopping someone from creating an ActivityPub server that federates with a Lemmy instance and spams upvotes from randomly-generated usernames. The server could just pretend that every username is a valid one.
Of course, I think something like that would be defederated pretty quickly.
Yes, there’s fediseer that can help with that sort of thing. However, I think if it wasn’t spam, it may not get noticed as quick as you’d think. Create a lemmy.world account, post a meme, and use your special server to throw 50 or 100 upvotes at it, and probably no one would realise.
The problem is that would also ban anyone else who might be on the same IP. VPN, Tor exit node, public WiFi etc.
Nothing beats a well staffed moderation team and other users willing to report abuse IMO.
Makes sense. Even then a lot of the troll’s comments are upvoted by a few people. There are definitely a few people here who share the troll’s extreme opinions. Thanks for doing your part in removing all these bullshit. I wish lemmy had a way to block account creation by IP addresses. That might stop trolls from making dozens of alt accounts.
We very often see the same username created across many instances, it’s very easy to do and Lemmy has no protections against it. Plus, there are no protections against creating multiple accounts to upvote your own posts (don’t get any ideas 😆). IP blocks wouldn’t work as instances are entirely independent, so there is no sharing of IP info across different instances.
Currently there is at least some level of coordination across instances, though, such as Lemmy.world’s Defense HQ, where instance admins can share info about spammers/trolls so we don’t have to wait for a report from one of our own users. There’s also Fediseer, but this protects against spam instances not spam accounts on mainstream instances.
Guilty as charged. I’ll say though, there are several legitimate reasons why one might want to do this. I personally use it as a substitute for Reddit’s multireddit feature, by grouping community subscriptions across different instances by theme. As long as users use the same username across instances I don’t think this practice should be automatically regarded as an attempt to sockpuppet. It that was the goal, the accounts would definitely not be using the same username across all the instances.
Personally I have accounts on multiple instances because I wanted to make communities in different languages and some instances focus a lot more on one language than others and also because the SFW instances defederated the NSFW ones. I do not really interact with the same posts though.
deleted by creator
I didn’t intend to imply that by using the same username across instances you were breaking some sort of rule. Different instances have different moderation policies, different federation policies, and different intents. Having multiple accounts in good faith should not be an issue and was not what I was trying to imply.
Rather, the intention was to show that we know bad actors do this with nefarious intent. Here’s an example (they show zero comments as they have been banned with content removed - also I think these ones only had posts not comments anyway):
Defense hq? Can I have a link might prove useful
Sure! It’s on the decentralised chat platform Matrix, which has heaps of lemmy stuff. If you need further instructions, feel free to ask 🙂
https://matrix.to/#/#defense:lemmy.world
Alright fanks m8
They don’t even have to be real accounts. Lemmy uses the ActivityPub protocol, and nothing’s stopping someone from creating an ActivityPub server that federates with a Lemmy instance and spams upvotes from randomly-generated usernames. The server could just pretend that every username is a valid one.
Of course, I think something like that would be defederated pretty quickly.
Yes, there’s fediseer that can help with that sort of thing. However, I think if it wasn’t spam, it may not get noticed as quick as you’d think. Create a lemmy.world account, post a meme, and use your special server to throw 50 or 100 upvotes at it, and probably no one would realise.
The problem is that would also ban anyone else who might be on the same IP. VPN, Tor exit node, public WiFi etc. Nothing beats a well staffed moderation team and other users willing to report abuse IMO.