Thanks for your reports on it. We try to action it as quickly as we can, and your reporting it helps us. We are attempting a few other Beehaw-only Lemmy changes to mitigate this sort of spam; short of de-federation. One of the major issues is that Kbin does not federate moderator actions. So they could have already dealt with the spammer and removed the posts, from Kbin. Those actions don’t get to us on Beehaw and the spammer posts still lives.
It’s a major hole with regards to spam, but in a hypothetical whistleblower scenario it makes it hard for a single entity to silence and remove all evidence.
There are simply trade offs, but it doesn’t mean there isn’t room for improvement.
Given the nature of the internet, I feel that spam is a much bigger problem than a potential whistleblower being silenced, and wouldn’t that kind of action show up in the logs anyways?
Can’t say we’re 100% responsible for the reduced spam lately; but wanted to get your thoughts. Has there been less Kbin spam lately? Do you feel our countermeasures are working well since a few weeks ago?
I remember a website that used a gamified interface that would pic a random picture uploaded by users and ask you to classify it. It was completely optional, accessible through a link and a great way to kill time.
Maybe a similar interface could be put in place, with proper warnings about the possible sensitivity of the posts you’ll see. The user sees the post and says if it’s spam or not spam. The randomness makes it somewhat tamper resistant.
I would gladly do it sometimes, instead of doom scrolling.
Thanks for your reports on it. We try to action it as quickly as we can, and your reporting it helps us. We are attempting a few other Beehaw-only Lemmy changes to mitigate this sort of spam; short of de-federation. One of the major issues is that Kbin does not federate moderator actions. So they could have already dealt with the spammer and removed the posts, from Kbin. Those actions don’t get to us on Beehaw and the spammer posts still lives.
OHHHH! This explains a lot. I had assumed kbin was doing nothing, but they could be removing the posts/accounts, but that doesn’t carry over to us.
Wow, that’s a HUGE drawback to how federation is handled, isn’t it?
It’s a major hole with regards to spam, but in a hypothetical whistleblower scenario it makes it hard for a single entity to silence and remove all evidence.
There are simply trade offs, but it doesn’t mean there isn’t room for improvement.
Given the nature of the internet, I feel that spam is a much bigger problem than a potential whistleblower being silenced, and wouldn’t that kind of action show up in the logs anyways?
As a former admin, I guarantee no one seriously reads those logs outside of happenstance. It takes the users speaking up to warrant that.
The whistleblower can read (or ask someone to read) the logs and find out who silenced them, then whistleblow that/get get whoever did it defederated.
Deleting spam isn’t optional. If you leave it there your community is dead.
What I meant with whistleblowing in terms of the fediverse is:
Whistleblower posts to instance A, and it gets mirrored on instance B.
Someone like Musk erases the post on instance A.
As the Fediverse currently works, the whistleblowing still exists on instance B, and cannot be deleted by an admin on instance A.
Asking a Musk to divulge who did the silencing is an exercise in futility.
That said, I’m totally on board with better tools to handle spam.
For real. I knew that Lemmy was a mess but this feels inexcusable. How could you NOT federate mod actions!?
It’s a bug which Ernest is working to correct.
It’s not a bug, and I don’t think it will be corrected. It’s fundamental to how federation works on Lemmy.
That is not what I have read Ernest say elsewhere.
Can’t say we’re 100% responsible for the reduced spam lately; but wanted to get your thoughts. Has there been less Kbin spam lately? Do you feel our countermeasures are working well since a few weeks ago?
I’ve been reporting the spam as much as I can, since I’m seeing a lot too, but that explains why it’s still hanging around. Thanks!
I remember a website that used a gamified interface that would pic a random picture uploaded by users and ask you to classify it. It was completely optional, accessible through a link and a great way to kill time.
Maybe a similar interface could be put in place, with proper warnings about the possible sensitivity of the posts you’ll see. The user sees the post and says if it’s spam or not spam. The randomness makes it somewhat tamper resistant.
I would gladly do it sometimes, instead of doom scrolling.
That would open us and the user up to legal ramifications of certain content.