On Oct 23, 2013, at 9:35 AM, Scott Brim <scott.brim@xxxxxxxxx> wrote: > On Tue, Oct 22, 2013 at 5:21 PM, Ted Lemon <Ted.Lemon@xxxxxxxxxxx> wrote: >> If the first message had been the beginning of a conversation, I would have taken it a lot differently, but the way it was dropped on the list with no discussion, it just looked like an attempt to start a massive flame war. That's exactly the sort of thing that the sergeant-at-arms is supposed to deal with. > I didn't mind the mail, but I didn't mind the request for suspension either because he/she had been sprinkling silliness across multiple lists, and content was becoming more disruptive. Seeing all the pushback from my earlier comment, I realize that there's a distinction here that we aren't making explicit, but that I think is important. It's inappropriate to spam IETF mailing lists. Spamming IETF mailing lists isn't covered under the warn first, sanction later model, because what you want with spam is for it to stop now, not later. To me, what's being described here is similar to situations we've seen in the past where someone subscribes to all the IETF mailing lists and starts posting conference invites on them. People have tended to be pretty patient about this, because it's not so utterly far off topic that it looks just like pill spam. But it really is spam—it's unsolicited, unwelcome, not contributing to a meaningful dialog. Randy says we should just delete it ourselves rather than blocking it. Given that people can't post without subscribing or being moderated, that's not completely unreasonable—the usual vector for spam is an unmoderated mailing list, and we don't have that particular problem. But we have in the past blocked people who post conference spam, and I thought that made sense at the time. I still think it makes sense. But I agree that in theory it doesn't match the warn-first policy. There are three ways we can deal with this, I think: - Keep the existing policy unchanged: in this case we should have sent a warning. - Add an AUP on subscription, and count that as a warning in the case of spam. - Change the policy to specifically say that spammers will be blocked without warning. My personal feeling is that the email that triggered this debate was a deliberate troll, and that trolls have the potential to be really disruptive. And so I think we _should_ block first and ask questions later, as a deliberate policy. We have appeals, and we have debates like this one, so I think that the harm of blocking too actively is minimal. But the various people who have argued that this is a new policy are right, and I think it's worth being clear on that.