-
-
Notifications
You must be signed in to change notification settings - Fork 884
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Instance banning a remote user should prevent them from participating in remote versions of communities #3399
Comments
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
wth 💀 |
Would that even work? I can't imagine the ban list is particularly big (since it's just text), and you'd need an unimaginably large amount of users (millions, or even billions?) to start taking up meaningful amounts of space. It's more likely that something else in Lemmy would break before then. At that point, it seems easier to just Federate a massive banspam list on a regular instance, without having to go to the trouble of bot accounts, and all of that. |
Well, we are not saying we should scrap the idea due to potential abuse. But it's definitely something we should keep in mind. Because right now, there is nothing that really prevents me from spamming millions of ban messages :P |
Instance bans are federated, but its possible that they are not going through in some cases. |
@Nutomic instance bans are only federeated into the modlog, they are not currently being checked when posting or commenting (unlike community bans). The point is that a user who is instance banned can freely post garbage in communities that are hosted in the instance that he is banned from, and this will ruin the experience for other users on his instance. |
This is the relevant code (which currently doesn't do much) for when a remote instance ban comes in: OTOH In the case of remote community bans, the bans are actually written into the local |
Ah you are talking about the user receives a site ban from an instance which is not their home instance. That is indeed not handled. Its a variation of #1205 |
Essentially a fix to this would be iterating over the communities of instance B, and for each one to run the code from the community ban handling, right? |
Yes that sounds about right. So you need to change the code inside |
additional related issue: i've noticed that unbans and restorations can federate in some cases, overriding a local removal/ban decision. while this certainly isn't working as intended, i think the security impact is fairly low, as such actions coming from other instances will typically lead to defederation, and similar actions could be performed by just creating new users, but i can see this, especially for the site ban, happen easily by accident, when a user gets temp banned on their home instance at any later time for likely unrelated reasons. |
Seems to be a dupe of #4118 , which I have a PR out for now. |
If I understand it correctly, #4464 does not address this for communities created after a site ban was issued. |
@Nothing4You you are correct - I will reopen this issue for now as I am working on a PR for the fix already |
Requirements
Is your proposal related to a problem?
Problem
When a user from instanceA is banned from instanceB, they are not prevented from making posts in instanceB communities. These posts do not show up outside instanceA, but this still means that future ban-worthy content from the banned user will be shown to all other users of instanceA, and the mod teams of instanceB will be powerless to improve the content for other users from instanceA.
Additionally, it seems that under certain conditions, these local-only posts from banned users can reach out to the wider federated network, for example, when a user from another instance searches for the post's url.
Describe the solution you'd like.
Solution
Instance bans would ideally work just like community bans - they would be federated and go into a new
site_person_ban
table, which could be checked everywhere that currently checkscommunity_person_ban
, and if a ban exists in either table, the users action would be prevented.Describe alternatives you've considered.
N/A
Additional context
No response
The text was updated successfully, but these errors were encountered: