The hidden cost of trolls…

One of our recurring trolls, the rather infamous Roger Maynard made his holiday visit to the Maple Leafs list to wish us all well.

So another round of locking the troll out goes on. We’ve found Roger’s two subscribed addresses, and he’s tried to resubscribe four other times so far, so I’m pretty sure I have all of his current accounts off the list. Of course, there’s always This, or perhaps this, from our repeat troll over on the colorado avalanche list.

Between the two of them, I have two seriously damaged lists, I know I’ve lost over a dozen users, and a number of key contributors who just got tired of the abuse and stopped contributing. And now, I get to waste time arguing with Roger yet again, time that could be used constructively, but won’t be, because I have to protect my system from someone who gets his jollies destroying things other people build, people who contribute nothing, but attack those who do.

And while I’ll win this battle, again, for a while, eventually, the history is he’ll be back later, and it’ll start again. And every time it happens, I get that much closer to saying the hell with it, knowing full well the day I do is the day roger finally wins and gets to destroy the list.

But this isn’t a rant, or a whine for sympathy. it’s a serious question for people designing social systems and any kind of software where groups of people gather and interact:

How does your system deal with the destructive troll? How does it give administrators and owners the power to stop their activities? How do you handle an abusive user who wants your system dead and your users scattered? Who won’t take “go away” for an answer, and who lives in a world where the person who answers the abuse address (if they answer it at all) tells you to go to hell? Who sees no problem creating dozens of yahoo and hotmail addresses to sneak back onto your systems and continue their attacks?

The blog’s comment spammers are well-behaved compared to some of these people. How do your systems deal with this kind of rare, but far from unprecedented behavior? How do you protect your users, keep your admins from giving up in disgust, keep your systems operating?

Or do you just assume that when you say “go away” they do?

Trolls this motivated are (fortunately) rare, but far from unknown. Remember this? or This? or maybe this?

So, how does your software handle the situation where someone doesn’t just want to raise your blood pressure, they want you dead? And don’t particularly care how they do it?

Related posts:

1. Three Tales of Trolls Derek Powazek – Three Tales of Trolls: Sometimes things...
2. On Trolls… Misbehaving.net has started attracting trolls, which should surprise anyone...
3. Usenet Trolls, by Gilbert and Sullivan (sort of) Special thanks to Derek DeLash, although I’m not sure...
4. word dilution jluster.org’s webvergnügen » 2 SPAM || ! 2 SPAM:...