no, think about what that would imply. Anyone could practicaly delete any other profile by creating an instance and banning that profile on that instance. That would be insane.
this post was submitted on 21 Jul 2023
16 points (94.4% liked)
Lemmy
12572 readers
3 users here now
Everything about Lemmy; bugs, gripes, praises, and advocacy.
For discussion about the lemmy.ml instance, go to !meta@lemmy.ml.
founded 4 years ago
MODERATORS
Having to play whack-a-mole banning on a per-instance basis if the account's home instance admin isn't cooperative isn't exactly ideal, either.
Perhaps what we need is to implement some sort of happy medium, such as having instance bans be "votes" towards some threshold that, if met, would result in banning the account network-wide?
But if you mean whether these posts are also deleted for people viewing your instance through another, federated instance, then yes, that is the case.
If you want to get rid of spam bots, ban them on your own instance and defederate any instance that doesn't do the same.
Today some spam bots posted about 20 or so posts, and they originated from thegarden.land, feddit.nu, lemmy.film, lemmus.org, feddit.nl, pricefield.org, lemmy.world, and geddit.social. I'm not sure I'm ready to defederate lemmy.world over a single account. There must be a better way.
Being able to ban people across instances would defeat the entire point of the fediverse. The best solution would either be to just defederate from instances that spawn lots of bots, or to contact those instances and work on the issue with them. There isn't really any other better way that I can think of
I have my own instance where I'm the only user and only admin. I have only banned two users (porn spam bots), but my ban list has hundreds of users, and I have no idea how they got here. I can see who banned them in the modlog and why, but how did they get banned on my instance?
Are they banned from a community that you follow?
I haven't checked that specifically, but that seems plausible. Do users banned from communities get federated with the community?
Yes, bans on a community are federated, so your copy of the community also removes their content.
User bans do federate. https://github.com/LemmyNet/lemmy/issues/1298
Yes, but only from the original community moderators / instance admins the community is on. This happens so a mod removing content also removes that content on any other instance following this community.
If you ban a user on your own instance, but originating somewhere else this will only make posts and comments by this user invisible for users on your instance.
how can a user comment from another instance?
my user on lemmy.ml cannot login on other instances, or can it?
As far as I understand it, the instance that hosts the user and the content is the only instance that can propagate deletions of the content/user. So if that user's home instance bans them, then the ban propagates. Otherwise, the ban is only local.
So is the right process to report the post? (I'm assuming reports go to the home instance's community moderator.)
And if that's the case, does that mean we need to avoid clicking the checkmark to "resolve" the report on our instance until the remote moderator has done something about the problem?
I'm actually not clear on how the reports work, but my assumption is that they go to both the instance you're reporting from and to the instance that hosts the content. I would also assume that the resolve status is local to each instance so it shouldn't matter when/if you click the resolve button.
But these are all assumptions, I don't actually know how the reports work.
Do you have some time to test it? You can post something to Ninja Tea Room and then report it. Let me know and I'll confirm that the report exists on our site. Then you can resolve the report and I can confirm one way or the other whether the status changes locally here, and whether the reason propagates. We will take care not to ban you. :-)
Sure! I just posted and reported it. Let me know if you see the report in your moderation queue.
The report just arrived in our moderation queue. Please click the resolved checkmark on your side.
Next, please post a second post in the Tea Room and this time I will report it, get confirmation from you, and then I will resolve it.
Ok, I have marked my report as resolved on my instance and took no action (so I have left the post up on my instance).
I have just made a 2nd post. I have not reported it.
Resolving the report on your side seems to have had no effect on our side. In Lemmy UI, there is a very subtle color difference in the checkmark before and after it is clicked, so I took a screenshot of the report.
Before:
After:
I guess there's a chance that it may take time for the change in state to propagate to me, so I'll watch it over the next 20 minutes or so to see if it changes.
I am now reporting the second test post.
Not seeing a report on the second test post yet.
It's definitely there. I guess this proves that some propagation is slow. This is what we see on our reports list right now:
Still not seeing it.
My guess is that actually the report doesn't get federated back out. I think reports are only federated to the instance that hosts the content and the instance the reporter belongs to. So in this case, lemmy.ninja is both the instance that hosts the content and the instance to which you belong to.
In the opposite direction, when I reported the first post, it went to my instance since I am the reporter. And to your instance, since your instance hosts the content.
Assuming I am correct, this could end up being a bit of a problem. That means, users on my instance could go about spamming the fediverse, and I would never see reports of their activity unless they are spamming communities on my instance. The only way I have to know that they're being bad users is if I notice we get defederated, if an admin of another instance specifically reaches out to me, if another user on my instance reports them, or if I manually monitor my users.
At a minimum, I would like to be able to click on a user's name, and be able to report them to their home instance.
But ideally, I would like to get copies of reports made on my users, regardless of where the content resides or who reported it, so I could swiftly take action on them if they're being bad users.
Assuming I am correct, this could end up being a bit of a problem. That means, users on my instance could go about spamming the fediverse, and I would never see reports of their activity unless they are spamming communities on my instance. The only way I have to know that they’re being bad users is if I notice we get defederated, if an admin of another instance specifically reaches out to me, if another user on my instance reports them, or if I manually monitor my users.
This is actually consistent with something that happened to us in the early days of lemmy.ninja. We had a few thousand bot accounts get created on our site. Some other sites defederated from us, but it took us weeks to notice that this happened. One of them happened to be a Mastodon instance, and that person indicated a ban reason that indicated that a user was an edgelord. Well, this was back in the beginning of our site, so we knew all of our users personally. If we had not been really on top of things and really plugged in to what was happening across a lot of the Lemmy instances, we would never have known that the bot users interacted with anyone. We still don't know how many posts or comments they made before we deleted them all.
I think you are right on all counts. I have just resolved both reports that we had on our side. That clears the report list, so I don't expect you to see any changes on your side, but I thought I'd mention it just in case.
Glad we got to the bottom of this! And yeah, just to confirm, I see no changes on my side.
However, I'm surprised to still see my posts there. I would have thought you deleting them on your instance would propagate out to my instance.
However, I’m surprised to still see my posts there. I would have thought you deleting them on your instance would propagate out to my instance.
Well, I purged them from the database. Maybe if I had removed the post instead of purging, that would have propagated. Right now the posts don't exist in our database at all.
But I bet the more likely scenario is that once a post gets propagated, it persists forever on the instance it gets propagated to unless someone purges it there.
Ahh, I'm actually guessing it's because you purged it. I think purges are local, but deleting it is considered a moderation action and is therefore propagated out. I just checked the modlog and clicked on a random post that was removed on a different instance. When I went to the post, I could still see it, but there was a notification at the top that said "Removed by Moderator".
Excellent! This is good to know! I bet other users who aren't admins/mods can't see the post at all when it's removed. I'll avoid purging from now on.
I bet other users who aren't admins/mods can't see the post at all when it's removed.
Yup, I'm pretty sure that's how this works. I opened the post in an incognito window, but still on my instance, and got an error message instead of the post.
Makes sense, you can open an issue for this if there isnt already.
Will do. By the way, thanks for all the hard work you've put into Lemmy comrade!
As I understand it, the account is registered on a given instance, and then that account interacts with other instances via the instance theyre registered on.
They cannot log into another instance with the same details (they need to create an account on each instance).
IE, to interact with you on .ml i’m logged into .world. The instances are federated so I see your post on this .ml community on .world
If I was banned from .world I can’t access the instance and thus any .ml because my interaction is via world instance. I can’t use the same account to log onto ml, I would need to make a new account there
view more: next ›