So I decided to get on Mastodon. This was one of those “I should take a look at what’s going on here” moments for me; I don’t regularly just willingly add myself to a new social media platform (I have *plenty* already) but I figured, this is a good time to start looking at what the buzz has been about amongst my friends and occasional enemies.
I also explicitly called out Mastodon as being a little unapproachable from a social friction coefficient perspective in my recent talk, The Dark Side of Free Software Communities, where I stated it has a fairly high social friction coefficient compared to Twitter or Facebook.
I figured I should put my money where my mouth is and actually join mastodon. Devote enough energy into actually overcoming these coefficients and I might find something interesting.
I should explain, a little, first.
In physics, there exists a “coefficient of friction” — how much force it takes to move an object along a plane or in reference to another object. A round ball on a flat plane has a low coefficient, whereas a dense rubber block will have a high one. This is vastly oversimplified, but you get the idea.
The “social friction coefficient” is how much effort (socially) it takes to maintain your peer group in a new social network. If all your friends are there already, or a vast majority of them, the frictional coefficient approaches zero. Conversely, if none of your friends are there or none of the content there is considered interesting to you, the coefficient approaches infinity. Compounded on this is a further scalar that indicates how friendly/approachable vs. “toxic”/unapproachable the community you’re joining is: If you only saw assholes on Facebook, would you join? (This is, incidentally, one of the compounding factors that makes Twitter frustrating for people.)
The “Technical friction coefficient” is how hard it is, at a technical level, to engage in a new social network. This includes, but is not limited to:
- Learning new terminology
- navigating a new application
- knowing if you’re doing the right thing
- Picking among options of varying quality
- Contributed documentation
Facebook hit a nice, prime spot:
- Memorable, monolith name: there is but one facebook dot com.
- They explicitly use well-known, friendly language: your friends like the posts you make to your wall — analogies that make sense to anyone who’s been in a college dorm or community shared space.
- The social graph encouraged “bring your friends” as well as “find your friends”. It made it very easy to buy in fast.
Facebook has a near-zero social coefficient and a near-zero technical coefficient.
You’ll notice I haven’t touched on surveillance. That’s intentional: Those who care about surveillance have an additional vector that has to be considered: The privacy coefficient. For the vast majority of humanity in a modern population, this coefficient is effectively epsilon. That, or it actively lowers one of the other two coefficients. This is the dangerous part: Facebook’s surveillance is pitched as a means to find your friends and connect you. It’s literally a selling point of the platform. The same applies to Twitter: The social graph that Twitter has for its users is stupidly strong, and it helps people find those who are aligned with their own interests, lowering the social coefficient.
So, let’s talk about my journey in the last few days of getting into Mastodon
Day 1: exploring
I found my way to Joinmastodon, where I got kinda overwhelmed. There was a lot of choice, and I quickly found myself going “… but which one do I choose?”
oh boy; let’s run through the thoughts that go through my head as a technically savvy person: Uhhhh… There’s some uhh… Japanese places? There’s one with the weird xyz TLD… I’ve heard of Framasoft before, aren’t they a company? Aren’t I supposed to shy away from the whole “coprorate ownership” thing? I don’t speak French, or Japanese, so uhhhh… Okay, so let’s try narrowing this down uhhhh
oh man, furries that might be fun but what’s chaos.social let’s go look at them.
(Also, if you’re wondering, yes, that is a rooster for the MSPSocial.net: It’s a local landmark; image my own)
What’s the first thing I see on vulpine.club?
Am I going to get in trouble opening this at the coffee shop? My parents house?
Right. Okay. Uhhh. Okay, This is… not quite what I expected — I mean, I like furries and all but this feels a little bit exclusive? Let’s try chaos.social:
Okay. I uh. Might be able to get into this? maybe? I’m not sure. Hmm… it says “small” — I… don’t feel like I’d fit in here. Uhh…
I’m going to give this a bit and come back tomorrow.
Day 2: A search engine is found.
this looks promising!
I find instaces.social through some exploration after stumbling upon some closed instances. The idea of a “closed” instance isn’t well explained, but it kinda makes sense to me? With nobody to really explain this, it feels weird.
So I tell it what language I speak and I’m asked how many users an instance would have? Okay, with no context I’d prefer having fewer neighbors, I guess.
Woah now, what’s all this about? Fuck it I don’t care. I just want to talk with my friends. Clicking “Next” gives me a blank list. The fuck? Let’s try the advanced button:
I can guess what some of these numbers mean! I want the one with the most people. Also, the thing saying “down” or not seems to be broken because those are all up.
Oh. Uh… down isn’t typically a good word. Let’s uh… try a.. oh, this seems broken:
okay, so Pawoo seems like a big one. Let’s look at that
Thirty percent of all users are on Pawoo.
Let’s look at their front page
So it’s
- in Japanese
- Filled to the gills with porn
Well, you know what they say…
If you dig for a little while longer, you’ll find that there’ a not insignificant number of places that “silence” Pawoo because of, well… Japan’s got a different way of viewing, socially, specific kinds of artwork.
I’ll leave it at that.
In fact, out of the top 10 largest instances by user count:
- The top is Pawoo, which is full of porn and blocked by a lot of instances
- Four of them are Japanese, which would be great if I spoke Japanese.
- Three are run by Pixiv
- Two are focused on porn and not run by Pixiv (Switter and Switter’s test instance)
- An overwhelming number of the rest of the content on the rest is in French, which would be great if I spoke French
- There’s a good chance you’ll see something very quickly on the front page of one of these instances with a post with… possibly objectionable content. More than a few times I encountered a good number of “yup that’s a Nazi” posts.
Day 3: I pick one.
I eventually picked mstdn.io — The guy who runs it is pretty clear what he expects out of the people on his instance, and he’s got a good community going. I was also just fucking tired of reading through codes of conduct here and there and everywhere.
The semi-official mstdn.io logo
I go through the hoops. I kinda know what I’m getting into already: I need to hunt down people. There’s a bridge that helps me find some of my twitter friends, but a large number of them seem to have vanished.
Then, as I join today, one of the other large instances, witches.town, is going to topple because there’s drama.
Takeaway
There’s a lot of Mastodon instances.
There’s a lot of really oddly specific ones.
If we consider the OECD findings on computer literacy accurate, Mastodon is inaccessible to all but maybe 25% of the population. It has a high social friction coefficient and a disturbingly exponential technical friction coefficient.
There’s nearly a dozen apps for Mastodon, all of varying quality. There’s no one “official” Mastodon Android or iOS client. There’s a list of suggested ones, but you have to dig around in a wiki on github to find that out. I found it once.
Overall, Mastodon needs to find how to get the coefficient of social friction down. A search engine for pepole might be helpful; A lot of my friends on Twitter are friends I found because I was like “oh this person I just followed also follows <these people> mutually. I’ll follow some of them too because we share interests”. It’s going to take a lot of organic growth.
A feature request I’d like to submit (and I plan on seeing I can hack on this) is being able to filter specific instances out of my view of the federated feed. As much as I like the federated stream, I’d appreciate the ability to Somewhat tune out certain instances for the overwhelming amount of content in a language I don’t speak or just don’t want to see for whatever reason.
Follow me on Mastodon: @indrora@mstdn.io