Open Thinkering


Tag: Fediverse

How to CW on the Fediverse

Preview of CW

This is just a short post that I can link to when people ask me a question about Content Warnings (CWs) on the Fediverse.

A good starting point is the 2018 Mastodon quick start guide:

My advice is simple: if you’re not sure whether a toot needs a CW or not, give it a CW. People really appreciate it and it doesn’t do any harm to be too cautious and too respectful of others.

You can also use a CW to summarise a long post. Some use it for joke punchlines. Maybe you’ll think of other uses for it. Have fun.

Things that I appreciate people putting a CW on:

  • Politics
  • War
  • Anything related to abuse
  • Strong eye contact
  • Spoilers

The way that I summarise it for people who say “why should I have to CW my posts?” is to say “so that people don’t unfollow you”.

What does this button do? The perils of being the other side of a screen to an algorithm

One of the things that has surprised me most about people migrating from Twitter to Fediverse platforms such as Mastodon has been people’s confusion over the ‘favourite’ button. It’s a star symbol and, as with in the earlier days of Twitter, it has no real function other than to the person pressing the button and the person who’s post is being favourited.

On Twitter, of course, the ‘like’ button (which morphed into a heart over time) had an algorithmic function; other people began to be shown posts that were liked, but not necessarily retweeted, by others in their network. Or by influential people. Or because its content was likely to provoke a reaction in you. Or maybe… something else? Who knows, the algorithms involved are proprietary.

In The Social Dilemma, there’s a powerful scene where a representation of a teenage boy is being created by the data he willingly hands to Big Tech companies. The algorithms are anthropomorphised in a way that helps less technical people understand what’s going on. While some of us may smile and comment on how it doesn’t really happen like that (algorithms don’t have intentions, as such) it’s nevertheless a good way of driving home the message that human flourishing is often at odds with shareholder value. Your ‘engagement’ with a platform might leave you angry for hours, days, or weeks just for a small increase in share price.

Meanwhile, on most Fediverse platforms, including Mastodon, there are essentially four ways of interacting with other people’s posts:

  • Reply
  • Boost
  • Favourite
  • Bookmark

There’s no ‘quote tweet’-like functionality on Mastodon, which is a purposeful attempt to stop sealioning. That’s why full-text search isn’t available, either.

These are all technical things, whereas what makes the Fediverse the Fediverse is the culture that’s been created over the last half-decade. As Clive Thompson writes:

Perhaps even more important than the design of Mastodon is the behavior established by its existing user base — i.e. the folks who’ve been using it for the last six years. Those people have established what is, in many ways, an antiviral culture. They push back at features and behaviors that are promoting virality, and they embrace things that add friction to the experience. They prefer slowness to speediness.


This is in part because Mastodon’s earliest communities included many subaltern groups who wanted to avoid the dogpiling harassment they’d received on major social networks — and understood that well-engineered friction would help.

What’s interesting to me is the culture clash between the way that the technology and culture works/worked on Twitter, and how it is in the Fediverse. It’s like the way that the former works is the ‘default’ in some way, and then that framing views how the Fediverse is seen.

I get it. I think I first wrote about being on Mastodon in… 2017? But you can’t look at something that is new to you and think about what it lacks compared to what you’ve grown accustomed. Not everything involves faster horses.

What I’m hoping that people are discovering is the peace of mind in not being encouraged by an opaque algorithm to make declarative statements in the hope for more engagement and virality. There’s something a lot more human, a lot more calming and convivial, about pressing a button and having a human on the other side.

On the importance of Fediverse server rules

Let’s say you’re going to set up a new WhatsApp, Signal, or Telegram chat for a group of people. It could be for family members, it could be for friends, or colleagues. It could be, which is a very familiar scenario for Team Belshaw, the way in which you find out when and where your kids sports matches are.

These chats have ‘admins’, people who have the power to change things like:

  • Group avatar/icon
  • Group name
  • Group membership (i.e. adding / removing individuals)
  • Deleting messages
  • Setting whether messages disappear after a set time

Perhaps most of the time, there’s no problem. But things can go spectacularly wrong, as I’m sure you’ve either experienced yourself, or heard about from others.

The above example is of people who already have something in common, or some kind of relationship that precedes the setting-up of the group chat. While I can imagine the manager of a sports team mentioning that participation in the chat is subject to the club’s privacy policy and code of conduct, you wouldn’t really do that for chat groups for friends and family. It would be weird.

Now let’s talk about places which are set up for conversation where the context is different:

  • Most people don’t know one another
  • They can talk among themselves, but also with people outside the group
  • When they talk with people outside of the group, the carry some of the group’s identity

Yes, I’m talking about the Fediverse. And, specifically, I’m talking about codes of conducts and ‘server rules’. Just as contracts are usually there to be referred to when things go wrong, so the server rules are there for when something goes awry.

If you’re a straight, middle-aged, white guy (like me!) playing life on the easiest difficulty setting, it might seem annoying to have to come up with server rules when you just want to set up a new Fediverse instance. Can’t everyone just be cool and get along? Well, frankly, no.

Conflict in social situations is inevitable; it’s the way that you handle each incident that matters. If you run a Fediverse instance with essentially no (or very few) rules, people playing life on harder difficulty settings won’t join. Moreover, some other servers might proactively block your instance. And even if they don’t, any small infraction from any of the people on your instance will lead to a look at the server rules. If you don’t have any/many, it’s likely to get blocked.

Most good codes of conduct are Creative Commons licensed. That way, we can built upon one another’s work. The one for can be found here, and exists largely thanks to the work of people more experienced than me. Like the Mastodon Server Covenant it’s not perfect, but provides a base layer for building a code of conduct that works for communities.

So if you’re setting up a Fediverse instance, be as intentional about the server rules as you are about the technology choices you make! Think about the behaviours you want to encourage. Read and learn those written by those running successful instances. Endeavour to create a moderation team with documented workflows as soon as you can. That way, it won’t be just people like me who feel safe and included — it’ll be everyone!