People seem to be talking about self-promotion on Mastodon. I should note that I'm only seeing the backsplash from it! No post has yet come across my feed saying "Do not promote commercial stuff on Mastodon."

This is probably because I came in through gamedev, and we launch and promote weird stuff all the time. "I put a game on Steam, it costs a few bucks!" is as indie as it gets in the gamedev world.

(Itch.io is also as indie as it gets, but in a different way. Separate topic.)

Anyway, I just want to say how weirdly familiar this all is from Usenet in maybe 1994. There was definitely a point where "everybody thought" that Usenet was definitionally non-commercial -- any kind of self-promotion of product or service with met with shock. Especially the infamous Green Card event (the first commercial spam message on the Internet).

Except... that attitude shifted. It couldn't not shift. Usenet was a primordial social network and platform (not the first of course, but one of the first) and it was growing. (Cf "the September that never ended".) More people means people doing more kinds of things. Some of them involve money.

Thus, says the old person in the most obvious way possible, history repeats itself. Sorry I don't have an exciting moral here.

Now, Usenet was structurally not at all suited for the kind of growth that 1994 brought. Notably, Usenet had only two moderation modes: "totally unmoderated" and "every post must be approved before it goes up." Neither of those actually work on the Internet-as-we-know-it, and Usenet mostly fell apart over the next decade.

(Many Usenet groups still exist, but they managed it by becoming so isolated and self-limited that nobody is left who knows to harass them any more.)

There's something of a parallel with Twitter there, by the way. The early years of Twitter felt like the early years of Usenet. Interesting people everywhere; interesting conversation just fell into your lap! And it was similarly not designed for the kind of growth that hit it. Its strengths in amplifying discussion turned out to be structural failures that amplified noise and harassment.

Mastodon's design, as I understand it, is a direct reaction to those failures.

@zarfeblong It is interesting because when Twitter first hit its big moderation problems I thought the solution would be to introduce new anti-harassment features. If as soon as Gamergate happened Twitter introduced dogpile detection and had, say, a way to identify "all these people messaging you rudely follow X. block all X followers for one week?" maybe harassment wouldn't have been so effective. But you had to individual click twice to block each harasser in the wave so it was ultra effective

@zarfeblong The reason this is interesting to me is because *Mastodon didn't take that path*. Mastodon has two anti-harassment primitives Twitter doesn't (block an instance, temp-mute) and one high-level moderation primitive (instances blocking instances) but otherwise the notification filtering looks like Twitter. Mastodon's approach, rather than staunching harassment, is trying to reduce user capabilities, even potentially useful ones, to make it harder to start a harassment mob to start with.

@mcc @zarfeblong

It's interesting to see which of these additional features would need to be features of Mastodon(/Pleroma/Akkoma/Honk/...) and which would necessarily need to be features of ActivityPub (or are impossible for any protocol of the same shape).

In particular, "block all followers of X for the next week" is trivially defeated by X making their followers list private. However, one could do something like "block all people who have mentioned X in their publicly-readable posts within last N days for a week". This is obviously a heuristic that relies on those actually being public, so relies a bit on secrecy. I'd really like to see exploration of what can (and what cannot) be done that doesn't rely on secrecy of mechanism.

@robryk @zarfeblong Getting well away from Mr. Plotkin's point here but:

Interestingly, when gamergate first started the harassment was mostly *not* being organized on Twitter itself. It was being organized on 4chan-style boards. However, those sockpuppets *did* tend to follow each other in a network, because of another site feature: Twitter limited, or was believed to limit, accounts that have few following/follower counts. Harassers made their accounts follow each other to avoid bot detection

@robryk @zarfeblong So you could imagine something like this working on Mastodon; combine network-seeking moderation with a quality metric that blocks signs an instance is hiding some follows. It might even be easier on Mastodon because the instances themselves are a signal.

This comes to a point I think I actually is important: You don't necessarily need to make harassment impossible, but if you simply add enough friction, you defeat very many— maybe even nearly all— cases of it.