+- Jellyfin Forum (https://forum.jellyfin.org)
+-- Forum: Announcements (https://forum.jellyfin.org/f-announcements)
+--- Forum: Forum Announcements (https://forum.jellyfin.org/f-forum-announcements)
+--- Thread: Recent spam and changes (/t-recent-spam-and-changes)
Recent spam and changes - joshuaboniface - 2023-11-13
Hello everyone.
As you may have noticed, the Jellyfin forums are under a major spam attack for the past 2-3 weeks. The spam is consistent (Korean text) and appears, based on the response to our mitigations, to be by actual human operators who have so far been able to elude all mitigations we've put up. Even remote spam blocking services have only been marginally effective.
To combat this, we've been trying a variety of techniques, some of which may seem very inconvenient. For this we apologize, but we must continue to explore all avenues to stop this and reduce the massive burden on our (like everything, volunteer) moderation team.
So far, the following mitigations have been put into place. Not all are guaranteed to be permanent.
1. A new security question, asking a less obvious question. This may change further in the future so ensure you read it carefully.
2. hCaptcha instead of ReCaptcha, which will ask everyone to complete two challenges.
3. A "you must post in a controlled way first" system with autopromotion of users. Please see the banner message at the top of the forum index for instructions when you first register!
4. Two external anti-spam registries: StopForumSpam.com and, as of today, CleanTalk.org. The latter MAY block VPNs and other such services as suspicious.
We are also considering our nuclear hammer - which I am strongly considering at this point - which is banning the entire country involved, but we are hoping the other applied mitigations are effective.
While we recognize that this is harmful to some of our community, we must balance that against the ability of our volunteers to handle the situation and prevent this forum from being overrun by spam. We hope you understand, and if you do run into trouble please don't hesitate to reach out to us on Matrix or via email.
RE: Recent spam and changes - joshuaboniface - 2023-11-13
We have, thankfully, finally come up with a decent solution to this: autobanning based on some common subjects and message bodies.
While it hasn't been a problem so far, we have added a new section to rule 6 requiring that ALL thread subjects be in English. If you don't, you may find yourself inadvertently triggering a ban. Otherwise we are being intentionally vague about the triggers to avoid tipping off the spammers.
If you are hit accidentally, please use the Contact form and link us a Pastebin of exactly what you tried to post, and we can proceed from there. Generally though, following rule 6 will avoid any issues.
At this time, we have also reverted both the CleanTalk.org plugin (since it would have required payment) and the onerous "first post must be in the Introductions thread" requirements. All other mitigations will remain in place for the time being, as they are indeed stopping a nontrivial amount of bot spam (they just didn't help with this particular, clearly-Human-made spam).