From: Ryan Pavlik Date: 2003-08-21T05:11:04+09:00 Subject: Re: [OT] Trollassassin On Thu, 21 Aug 2003 04:11:29 +0900 "Hal E. Fulton" wrote: > ----- Original Message ----- > From: "Kurt M. Dresner" > To: "ruby-talk ML" > Sent: Wednesday, August 20, 2003 1:35 PM > Subject: [OT] Trollassassin > > > > So, I had this idea, but I couldn't think of anywhere better to post it. > > > > How about a recognized troll database that could be queried by something > > similar to spamassassin and tag mail as from a known troll? > > Interesting idea, but some practical problems. > > 1. Spam is identified largely by content, not necessarily > by sender. Spam is less personal. Hmm, it seems to me this would make it _easier_ to do trollassassin. > 2. Could possibly add a troll-rating value as a header... > then each client could select a threshold. > 3. Trollhood is in the eye of the beholder. We don't all > always agree. Both of these together and we have a solution, perhaps. You submit someone's address to trollassassin, and it gets voted on, and maybe something like Google's PageRank is used for making sure trolls don't "strike back" or "trollwhack" the database. > 4. Who decides who is a troll, and what keeps that system > from being abused? Can my top ten enemies get me labeled > as a troll? What about one single enemy with ten email > accounts? Voting would help here, and voting could go both ways (+/- trollness). Perhaps everyone would have their address entered in the DB (a valid address would be required to vote), and that might help combat abuse. I think I'm missing a few possibilities... but I like his idea. OTOH, I hope it doesn't turn into a Slashdot Moderation System for email addresses... ;-) -- Ryan Pavlik "You mean super bouncy ball!" - 8BT