@UlrikeHahn@fediscience.org @julian@fietkau.social
Exactly. The solution is obvious, there are many servers with similarly reliable track records, promote one of those.
This is something I continue to think about a lot. I thought changing the default was a bad decision at the time and haven't changed my mind! It's not just the centralization aspects of it; it's also that (based on retention rates) most people don't have a good experience on .social -- so they wind up leaving fedi.
Rotating the default doesn't seem to me like it would address the :"good experience" aspect of the problem. For most people who are looking for a Twitter-like experience, .social's as good an approximation as anywhere else in fedi -- not great, but other instances aren't any better. And for people who are looking for a local community that aligns with their interests or geography, they're not going to find it on other largeish open-registration instances (and it doesn't make sense to have anything but a largesish open-registration instance as the default).
@julian@fietkau.social
@FediTips@social.growyourown.services @UlrikeHahn@fediscience.org
@thenexusofprivacy@infosec.exchange @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org Yes, exactly on point. The best case scenario for the Fediverse is a rough alignment of servers with cohesive communities, because if the community matches the infrastructure, that's good for moderation, long-term stability, and the day-to-day experience of each individual. A default server can't provide that.
Ideally, everyone would join fedi by being invited to a well-moderated small-to-medium server by a friend.
Approaches that might work better involve integrating a good instance picker into the signup process, or an onboarding flow that treats the initial instance as a "starter instance", a base for exploring that makes it easy to move to another instance. Realistically though it's not clearly how likely it is that Mastodon gGmbH will prioritize the work that's needed to support either of these -- which isn't an argument against pushing for them, just that we should be looking for other alternatives as well.
In general it seems to me that might be better to focus our efforts in terms of making it easier to join communities in fedi as a whole, not just Mastodon. For many people something other than a Mastodon-based instance may well be a better choice. Of course that still leaves the problem of people who search for "mastodon", or have read an article about Mastodon and followed the links to either the Mastodon app or joinmastodon ... but I don't know how to address those without Mastodon gGmbH's cooperation,
@julian@fietkau.social @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org
@thenexusofprivacy@infosec.exchange @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org I'm personally also quite intrigued by the idea of treating the default as a “tutorial server”. Video games manage to design sandbox areas that anyone can use to try out new toys, but that naturally get boring after a while unless you leave them and explore. I think trying to translate that concept to fedi servers could be fruitful, albeit difficult.
And actually I'd go farther and say it might be better to focus on community-led alternatives to commercial social networks in general -- not just fedi. For people looking for a US-focused, Black-centric microblogging experience, Blacksky might well be the best option today (and as Northsky becomes more real, that's likely to be a good option for a North American-focused 2SLGBTQIA+-centric microblogging experience). For people looking for a Palestinian-friendly photo/video-sharing app, Upscrolled might be a good option even though it's not decentralized.
@julian@fietkau.social @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org
@thenexusofprivacy@infosec.exchange
I feel that its worth pointing out in these conversations that nobody starts new servers anymore. Its hard (to measure kinda impossible now since fedidb removed the 'first seen' statistic, and server age does not show up in nodeinfo), but virtually no new servers have been started since 2024 that have gotten over 1k mau, I think it was like 4 total (this spring when the stat was still available)
@julian@fietkau.social @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org
@julian@fietkau.social @thenexusofprivacy@infosec.exchange @FediTips@social.growyourown.services very much yes to the tutorial server idea (I think of it in terms of a paddling pool…) and the idea would very much be you get funnelled out if it once you’ve got a basic understanding of what choosing a server actually entails and to what extent it does (and doesn’t) matter.
(but if all we have is a default server then it would still be better they rotated, and default servers should also be exceptionally well moderated to qualify…)
@UlrikeHahn@fediscience.org' has also brought up the idea of something similar to a "tutorial server" with limited functionality ... it's intriguing. one challenge is how to get enough interesting stuff there for people to decide it's worth exploring further. If it's just a tutorial that people have to go through to get to the next stage, then the risk is that unless folks are really really committed they're likely to lose interest.
@julian@fietkau.social @FediTips@social.growyourown.services
It's a great point. Fediverse Observer still h as this info and from a quick pass it seems like the only new Mastodon instances with > 1K MAU that have started in the last two years are qlub, a québécois social network, and the alt-right server noauthority.social (which isn't completely new, since it was a split from noagenda.social). There are also a handful of peertube, wordpress, nodebb, and snac (!) instances.
( https://fediverse.observer/list and sort by Active Users)
@laurenshof@indieweb.social @julian@fietkau.social @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org
@laurenshof@indieweb.social @thenexusofprivacy@infosec.exchange @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org Anecdotally, I have been in contact with a professional association that is cautiously interested in hosting a Mastodon server for (possibly) a few hundred MAU in Germany, but is very scared of the social media liability laws that have gone into effect lately, some of them on pretty short notice.
Talking to them is an interesting mix of “we don't have a mandate to take this kind of risk” and “if we can't figure this out, then who can?”
imo a core problem is that perceptually, an app is usually tied software (the mastodon app) instead of network (the fediverse app) or server (the indiewebsocial app)
@thenexusofprivacy@infosec.exchange @julian@fietkau.social @FediTips@social.growyourown.services @UlrikeHahn@fediscience.org
@thenexusofprivacy@infosec.exchange @UlrikeHahn@fediscience.org @julian@fietkau.social
"Rotating the default doesn't seem to me like it would address the :"good experience" aspect of the problem."
It's not about that at all, it's not about UI or UX.
The problem is about what happens to the network if it centralises. There are many important reasons why the Fediverse is decentralised: https://fedi.tips/why-is-the-fediverse-on-so-many-separate-servers/
All these reason get squashed if mastodon.social becomes over 50% of the network. The network would eventually "enshittify".
@julian@fietkau.social @laurenshof@indieweb.social @thenexusofprivacy@infosec.exchange @FediTips@social.growyourown.services this is the kind of situation this new Mastodon service is aimed at, no? https://joinmastodon.org/hosting
We're looking at things differently.
From my perspective, .social's steadily increasing percentage of Mastodon users is only part of the onboarding problem. Any improvements are going to require either a significant change of attitude from Mastodon gGmbH or a completely different way of thinking about onboarding -- or both! So I'm more interested in trying to address the issues more broadly.
You're separating it into different problems, which is also a valid way of looking at it ... but if the problems are interetwined, then there's no guarantee that the effort put into making progress on one problem gets things noticeably closer to an overall solution. My guess is that we have different intuitions about how decomposable the overall problem is.
@FediTips@social.growyourown.services @UlrikeHahn@fediscience.org @julian@fietkau.social
@julian@fietkau.social @laurenshof@indieweb.social @thenexusofprivacy@infosec.exchange @FediTips@social.growyourown.services this is the kind of situation this new Mastodon service is aimed at, no? https://joinmastodon.org/hosting
We're looking at things differently.
From my perspective, .social's steadily increasing percentage of Mastodon users is only part of the onboarding problem. Any improvements are going to require either a significant change of attitude from Mastodon gGmbH or a completely different way of thinking about onboarding -- or both! So I'm more interested in trying to address the issues more broadly.
You're separating it into different problems, which is also a valid way of looking at it ... but if the problems are interetwined, then there's no guarantee that the effort put into making progress on one problem gets things noticeably closer to an overall solution. My guess is that we have different intuitions about how decomposable the overall problem is.
@FediTips@social.growyourown.services @UlrikeHahn@fediscience.org @julian@fietkau.social
Fedi.Tips
The Nexus of Privacy
