At a meta level, something I find mildly interesting is how many people are writing stuff on Mastodon about how it's impossible for Mastodon to scale up without using an ad supported model (b/c server costs), it's better to have ranked feeds because most people want them, etc.

The thing I think is interesting is that the people writing this stuff, implicitly, seemingly cannot conceive of a model where the organization is not growth and profit maximizing.

If I look at my own site, I make $30k/yr off donations and could easily do 10x that with big ads, but I don't currently need the money enough to put ads up.

Likewise, if you do the math on how much it costs to run a Masto server, even if the fediverse gets way bigger and hosting costs go up, you should be able to run a small instance on donations.

Donations aren't profit maximizing compared to have ranked feeds with ads, but it's ok for there to be things that aren't profit maximizing.

It's so obvious that it's ok for things to not be profit maximizing that it's sort of absurd to think that someone would think that someone would say that things must be profit maximizing, and I think that if you asked the people writing the comments I'm referring to, they'd agree in the abstract that it's ok to not maximize profit.

This is the really insidious thing about absorbing values from the environment around you.

@danluu it has to be sustaining tho, and I suspect that if you hadn't earned big money from your last jobs, you wouldn't be willing to live on 30K/year from your Patreon. Maybe you anticipate growing to a large enough number that it becomes sustaining (100K/year?).

Sustenance has to include paying a fair wage+benefits to the admins.

@heathborders I can't speak for the admins, but I lived on $12k/yr for a long time when I was making a lot more than that as an engineer I don't have any problem living on that much.

I think an instance needs to be quite large before it really needs a full-time admin. This isn't mastodon, but the lobsters admin is very part time and does a great job with that. And HN is huge (surely larger than the largest masto instance) and gets by decently with basically one full-time admin.

@danluu @heathborders Can you show your work on HN being bigger than the big Mastodon instances? That seems unlikely to me.

mastodon.social has hundreds of thousands of MAUs, so probably at least in the hundreds of thousands of posts per day (though it's hard to guess that ratio precisely.) That seems like far more than what HN handles.

@gregprice @heathborders I was thinking traffic and not posts or words written. I agree that Mastodon probably has more posts and more words written.

OTOH, I don't know what dang's split between programming and moderation is, so he might be much less than full time.

@danluu @heathborders Cool, thanks for the reply.

For thinking about the scale of the moderation burden, I'd expect the number of posts and of words written to be much more of a driver than the amount of read traffic. So I think that points to Mastodon (and in particular the largest instances) needing more moderation effort than HN.

@gregprice @heathborders Yeah, that's fair. One thing I wonder about is how good the moderation tooling is on HN vs. Mastodon.

My impression is that the moderation tooling is very good on HN. No idea about Mastodon, but from the non-moderation interfaces I can see, I'm going to guess that it's poor, which would really increase the mod burden.

@danluu @heathborders Yeah, that's a good point. I agree.

Another "unfair advantage" I think HN has on moderation is the fact that it's a forum for a specific, limited range of subject matter.

Lots and lots of the sorts of posts that tend to make content moderation hard can be just promptly squashed on HN as off-topic. On a general-purpose social media site, that won't wash.

@gregprice @heathborders A question I have is, in the long run, if the fediverse scales, is small instance moderation even possible for instances that don't want to basically defederate without outsourcing moderation to a small set of centralized actors?

I've seen a variety of solutions to this and none even seem remotely plausible to me, e.g., see https://www.facebook.com/dan.luu/posts/pfbid023LoPRovpoprimEhSUSbaCTxKXfzGBgANk2dLr2DwFdHpgUwhSZiyA5C2rLMakrW4l?comment_id=450688213862987 for a discussion where someone proposed some solutions that I think can't possibly work.