The discussion revolves around whether Discourse should implement “slowban” or “errorban” features—subtle, secret moderation tools that limit user activity without explicit notification. dfabulich references Jeff Atwood’s 2011 blog post describing hellban (shadowban), slowban, and errorban as effective, low-drama ways to deter problematic users. He argues that secret bans prevent retaliation and encourage voluntary departure. Falco counters that locking users to lower trust levels already achieves similar throttling effects, as seen on Meta. dfabulich emphasizes secrecy as key, noting that visible bans can provoke backlash. Falco clarifies he meant rate-limiting via trust levels, not full bans. dfabulich replies that trust-level restrictions are not secret, showing a screenshot of the visible rate-limit warning. sam, representing the Discourse team, firmly states there are no plans to build slowban, errorban, hellban, or similar retaliatory features into core or official plugins—these must remain third-party. Stephen humorously latches onto “disemvowling” as a favorite term. anon48433008 notes enterprise customers can’t use third-party plugins, suggesting potential friction between user desires and platform policy. justin offers a philosophical counterpoint: better moderation, communication, and giving users chances to improve are more dignified and effective than covert tools. The consensus leans toward transparency and human moderation over hidden punitive mechanics.
Back in 2011, Jeff wrote an excellent blog post discussing types of alternate bans:
A hellbanned user is invisible to all other users, but crucially, not himself. From their perspective, they are participating normally in the community but nobody ever responds to them. (Hellban is also known as “shadowban.”)
A slowbanned user has delays forcibly introduced into every page they visit.
An errorbanned user has errors inserted at random into pages they visit.
I see that there’s a Discourse Shadowban plugin. But shadowban/hellbans are fairly easy for the banned user to discover: just browse the forum logged out.
Would the Discourse team consider implementing slowban or errorban? Or would that have to be a plugin?
What a slowban does that locking a user to trust level 0 (and the consequent rate limits this brings) doesn’t?
IMO locking problematic users, who aren’t as bad to warrant a suspension or a ban, to lower trust levels and they rate limits works fine. It’s what we do on Meta.
Formal suspensions and bans can result in the user retaliating, even off-forum. “A ban! They’re trying to silence me! I’m getting a response and that’s exciting! I’ll try to rally the forum to show how evil the moderators are!”
Secret bans are a way to penalize a user without causing the user to retaliate. Ideally, they just lose interest in the forum.
As Jeff pointed out in the post:
I’ve personally talked to people in charge of large online communities – ones you probably participate in every day – and part of the reason those communities haven’t broken down into utter chaos by now is because they secretly hellban and slowban their most problematic users. These solutions do neatly solve the problem of getting troublesome users to “voluntarily” decide to leave a community with a minimum of drama. It’s hard to argue with techniques that are proven to work.
That is not what I suggested at all. I asked why not lock this user in a lower trust level in order to rate limit the user, effectively making him post slower (as in less post frequency).
We have no plans in the foreseeable future to build slowban, errorban, hellban, disemvowling, automated retaliatory DoS attacks, downloadfloodban, bitcoinban and any other retaliatory features into the core product.
We also have no plans to build plugins around this area. This kind of stuff would have to be in a third party plugin.
Just bear in mind customers don’t get that option, a potential exists for bitterness, might want to simmer back burner ideas how to get the two to meet, customer desires and workforce responsibilities,
I guess that’s why the limitations, extra work when things don’t mesh well
Another take is that more powerful banning tools answer the same problem as better moderation. While there will always be trolls, many cases of troublesome users can be corrected with consistent and communicative action from mods. A community is not a democracy (so thus enforce the rules of participation), but give people a chance to make better choices with their behavior first. It’s the dignified and respectful thing to do, in my opinion.