For your own sanity, you have to remember that not all problems can be solved. Not all problems can be solved, but all problems can be illuminated.
—Ursula Franklin, in an interview with Meredith Whittaker*
Before going on to the next item in the Ursula Franklin series I accidentally kicked off last week, I wanted to write a little short post about a very practical set of network decisions by the makers and users and refusers of Bridgy Fed. But little short posts expand in my brain like those tea flowers that bloom and expand when you put them in hot water. My brain is also in hot water right now. And I couldn’t leave Franklin alone after all.
Let’s begin with the bridge anyway, but if you want to skip down to the more general part, go for it.
Bridgy Fed—which for the rest of this post I will call “the bridge”—is a clever bit of connective tech that allows people on Bluesky to read posts on Mastodon and the rest of the fediverse, and vice-versa. You don’t need two accounts to make it work, you just turn it on (instructions here) from either your Bluesky or your fediverse account, and the bridge will create a mirror that people on the other network can follow and interact with. I do this with both my Bluesky and Mastodon accounts.
Why though
I have accounts on the fediverse and Bluesky and I use them differently, rather than cross-posting identical things to both. Most culture and political stuff goes on Bluesky because that’s where my culture and political people are concentrated, and because the reply controls there let me deal better with inevitable unpleasantness than Mastodon does. Most of my social internet thinking and discussion happens on the fediverse because that’s where my network nerds are concentrated, and because I love the fediverse model and try to give it most of my attention.
But I want to let people follow whichever set of posts they want, from either network, so I use the bridge.
Bridging also means that if I decide to abruptly get off Bluesky or my Mastodon server, or if they abruptly delete my account or whatever, people on the network I’m no longer “on” will be still able to find and follow me.
The very best candidates for bridging aren’t nerds like me: they’re the people who really want to be reaching a broad readership (or a narrow one spread across multiple networks) and who don’t want to deal with multiple accounts on multiple networks. Journalists, public health folks, elected officials (really their staffers), people who are just good at explaining important things, artists and artisans who want to bring their work to broader audiences could all be served well by activating the bridge for their Bluesky or fediverse accounts so that interested people can follow them from their home network.
Why not
Bridging broadens risk surface. I’m “on” both the fediverse and Bluesky already, so I’ve already made the decision to share my messages in both of those ecosystems. So for me, the main risk is that I lose a little bit of friction in controlling who sees my messages and what they get to attach to those messages in the form of replies (and quote-posts, where applicable).
If I post “I love pancakes” on Bluesky and my replies fill with people yelling about how I must hate waffles, I can block or mute the wafflers, hide their replies, throttle replies so only people on a specific list can reply, or turn replies (or quote posts) off completely. This is true for replies that originate on Bluesky and replies that come across the bridge from people on Mastodon. What I can’t control is what happens in the replies visible on my bridged post over on Mastodon—in that mirror that represents me, but that I don’t inhabit in the way I inhabit an account fully under my control.
So the question for me, bridging from Bluesky to Mastodon, is “How much do I care about what people on Mastodon who look at the replies on my mirror will see?” The answer for me right now is “Eh, not enough to outweigh the benefits.” If my threat model and risk-benefit calculations change, I might change my mind. The same calculation applies when bridging from the fediverse to Bluesky, except that posts I mark “unlisted/quiet public” or “followers only” (or “mentioned users only”) won’t get bridged to Bluesky at all, which is great.
If you’re on the fediverse and you are categorically opposed to Bluesky on the basis of vibes or principles, you’ll have to make your own calculation about whether it’s worth letting people on Bluesky follow your posts. Likewise, if you’re on Bluesky and can’t stand the fediverse, you’ll have to make the same calculation.
Who runs the bridge
A New Social is a new US-based nonprofit founded by Ryan Barrett and Anuj Ahooja that will provide governance and structure for network interoperability tooling including the Bridgy Fed network bridge. Ryan, who was mostly recently at carbon exchange company NCX, is the longtime maintainer of Bridgy, which connects websites to centralized social networks, and the developer of Bridgy Fed, which creates bridges between websites, the fediverse, and Bluesky. Anuj, most recently at Flipboard, joined up with Ryan to focus on decentralized interop. Here’s Anuj’s thinking about the move. I think this is good and important work, and that there are ways to do it that minimize disaster while providing plentiful benefits. More on that another time.
Wasn’t there a whole thing about the bridge?
There was a whole thing. When Ryan introduced his ideas for implementing Bridgy Fed, there was a storm of criticism from people who didn’t want to use it, or who didn’t want it to exist, or even who maybe did want to use it but thought making it opt-out was too invasive (that last group included me). It got heated and personal very quickly, in the way that these things do. Laurens Hof covered it on the Fediverse Report, and got to the underlying issues with admirable concision.
After the storm cranked up, Ryan landed on opt-in bridging, which dramatically reduces uptake, but works well with the ethos of consent that some factions of the fediverse treasure. I think that was the right call for the fediverse side of things.
The fight over how Bridgy Fed would work—and whether it should exist—was part of a running disagreement about what the fediverse is and how it should work, as Laurens notes in his post about the blowback.
[Thanks to Jennifer Moore at Uncharted Worlds, who wrote about the bridge at the time, for pointing out that I hadn’t described the storm and its outcome with quite enough precision, especially the fact that the bridge was proposed as an opt-out service. That’s clearer in the linked material, I think, but not clear enough in the post. I fixed it up. —ek]
The standing wave on the fediverse
A thing that happens like clockwork on the fediverse is that someone proposes or starts a new project, and the most privacy-minded factions of the fediverse shut it down by a combination of reasoned opposition, high-velocity namecalling, and trolling. This makes me think about what Seth Frey and Nathan Schneider call “affective voice,” as distinguished from “effective” voice.
The two standard personality-based explanations of this phenomenon go like this:
- Either some people on the fediverse are just entitled, toxic bullies opposed to anything that might change the network, whose norms should be derived only from what’s possible at the protocol level, or…
- Fedi developers are arrogant and entitled tech bros who run roughshod over vulnerable and marginalized people, and therefore deserve shaming and maybe sometimes (allegedly) having illegal and deplorable content remote-loaded onto their servers before being reported to the authorities.
And look: I don’t doubt that some privacy advocates are permanently truculent and looking for someone to throw bile at. I don’t doubt that some developers are callous or just not thinking the problems through. But I don’t see evidence that that’s the main or even most usual problem. I see instead a lack of clarity and a real dearth of ways for the fediverse’s various interested constituencies—let alone those of the wider networked public—to talk to each other in depth and come to common understandings before and outside of super-heated blow-ups. (Some may find the blow-ups energizing and cathartic, but I think the greater portion are exhausted.)
When you start getting into the details of personal and community threat models (varied! undocumented!) and cultural expectations and network philosophies, the terrain becomes profoundly complicated. People in various fedi communities—along with the amorphous group floating confusedly between—have conflicting needs, desires, and models of place and space. That the stakes are so high in our ridiculous historical moment makes talking harder.
I’ve been stewing on four or five rounds of the fediverse standing wave, and I think its current form has three major causes.
1. Conflicting models of what the fediverse “really” is
To oversimplify: Some people define fediverse connectivity according to the maximum breadth permitted by the underlying technical protocol, which permits universal connectivity between all ActivityPub-based services. Plenty of people would like to see unfettered connection beyond that point.
Others define it according to the consent-forward, privacy-conscious social protocols that emerged from the past decade or so and which have been largely shaped by:
- the fediverse’s relatively tiny scale,
- the often opaque and anti-findability affordances of fediverse services that have provided a variety of partial securities through obscurity and friction, and
- the outsider/marginalized character of many early fediverse-adopting communities, including LGBTQ folks, furries, anarchists, and sex workers.
I don’t think these two models can be made to agree, which means the fediverse is an ideal place to work toward a pluralism that allows multiple models of the network to co-exist—and help individual humans find their place within them. Island networks and closed-by-default implementations get us part of the way there, but I think it’s going to be necessary to more fully illuminate the terrain of our (currently conflicting) models so developers can understand how to design for multiple constituencies and cultural models.
Social threat modeling should be a significant part of this, and is a fantastic target for privacy-preserving cross-community collaboration efforts. Ryan called out the need for social threat modeling a year ago in his first Bridgy Fed post, and Jon Pincus has been writing and cross-linking about it since 2018.
2. No way to see potential threats coming
If you see the fediverse as an alternative to ultra-surveillant centralized platforms, or you rely on it as a way to communicate about things that would on other platforms attract overwhelming abuse (or over-moderation), the sudden appearance of new fediverse tools and services that appear to shove you and your communities into non-consensual indexing and “discoverability” is jarring and threatening.
There’s no built-in way for fediverse members to review and evaluate changes and additions to the ecosystem that might make things worse for them, and you can’t opt out of something you don’t know is happening. This feels bad!
And it means that if you don’t want to be connected to tools and services you find too risky (or just repulsive!) you have to maintain constant vigilance, and this takes the form of watching out for people posting that a new thing has arrived and it’s bad. Put another way, in the absence of technical affordances for maintaining boundaries, the immune system of the privacy-seeking and visibility-skeptical side of the fediverse is social.
So the fediverse is technically open and culturally anxious about it.
3. Few conventions for the exercise of effective voice
If the only way people can get the attention—or believe they can get the attention—of would-be fedi developers is by mobbing their GitHub issues with comments about how they’re horrible people, that is what will happen. (Riots being the language of the unheard, etc.)
The best time to understand fediverse cultures and model for their threats and experiences and engage in co-design work is before launching even a proposal. But we don’t really have great ways for communities that inhabit the fediverse and developers who want to build good and useful fediverse tooling to work together. I think this is very fixable.
Darius Kazemi’s request for community comment on a proposed tool for helping fedi developers play well with other fedi services is a great example. Darius a.) did a bunch of research and thinking to define an non-invasive way to proceed, b.) explained the risk landscape and his reasoning in detail, and c.) brought it all to the fediverse for a lengthy comment period. I think skipping any of those steps would have led to trouble.
Scrupling & scrambling
My work over these past couple of years has focused on illuminating the situation rather than valorizing or vilifying any given position on these issues, even when I strongly hold one myself. And I try to make that distinction! I would rather eat glass than talk to Threads, because of what I know about Meta. But I am extremely interested in helping people who want or need to connect to Threads to do so with as little damage to themselves and their networks as possible.
In the same way, I have come around to being pretty positive on bridging for people whose risk-benefit calculations make it good for them and their networks, but I am also extremely interested in helping people who hate the idea of bridging avoid being inadvertently sucked into a cross-network vortex they don’t want.
In Whittaker’s essay I linked at the top, she both quotes and synthesizes Franklin’s thoughts about “scrupling,” a Quaker term that appears throughout Franklin’s voluminous body of work:
“It comes out of the anti-slavery movement, originally. People would get together to ‘scruple,’ that is, discuss and debate a common problem, something they had scruples about—say, justice—for which they did not have a solution. This is scrupling, and this is something you and your friends can do.”
Gather and talk. Empathize and listen. Don’t chase the spotlight, and accept that some problems are big, and difficult, and that what you’re good at may not fix them.
In the interview, Franklin follows “Not all problems can be solved, but all problems can be illuminated,” with a bit of wisdom that made me laugh out loud: “If the eggs are scrambled, they’re scrambled. You can’t unscramble them. All you can possibly do is cook them and share them with somebody.”
I feel pretty scrambled. But I think illumination and scrupling are a good chunk of what I am trying to do here, with you.
Notes
* The essay is credited to “Meredith Meredith” but Whittaker claims it here.
Thank you so much to supporting members. This work feels so weird right now, but also so important, and I would not be able to do it without your backup and your company and your help. (You can join up here if you’re interested.)
The feature image at the top of this post is from the US National Park Service. The full caption for the photo reads, “Looking north across the Merrimack River in Lowell, MA as workers construct a temporary footbridge. The flood-damaged Centralville Bridge is at left in the photo. Although the footbridge was removed when the new Centralville Bridge was completed, the concrete base for the tower on the north riverbank is still visible today.” Its archive record is “PL&C BOX12_LOWE 9249 Copy2 NPS/Proprietors of Locks and Canals.”
I don’t know how long the extraordinary collections of historical images that NOAA and the NPS and so many other federal agencies maintain as a public good will remain visible. I’ve spent dozens of hours lost in these images over the years, tranced out over the monumental effort of care they depict and represent.
At the Covid Tracking Project, a principle I came back to over and over again in our wrangling with the controllers of federal public data was the data already belongs to us. It belongs to the people and always did.
Much more immediately important federal data is currently disappearing from view, including data I’ve worked with myself to try to save lives. So much of the infrastructure of our civic life in the US and beyond it is under existential threat. I don’t know why the looming loss of these unnecessary but beautiful gifts is the thing that finally made me cry.