Most of the servers are closed to registrations, and we are just talking about a couple million new users over 2 months. What if 100 mil users want to move?
Are we going to relive the Twitter history, which crashed every time Justin Bieber tweeted?
If you're not willing to pay for the services you require... you're going to have a bad time. That is the ultimate lesson of the Fediverse.
On the technical side, the whole thing is lossy as hell. If I view a foreign profile on my instance it shows a handful of posts and then says that "older posts" aren't shown from other instances. Then if I go to the other instance there are hundreds of posts that aren't older, they were just missing on my instance. None of it really works or makes any sense.
Didn't you just describe Twitter?
Of course they do. If the instance administrators do not abide by the rules they set, every user has the possibility to jump ship to an instance whose team actually does its job properly. This is undoable on Twitter, which makes me doubt that the overly general statement of
> The entire model is strictly worse for the user than Twitter was.
holds any actual merit.
The model does make it difficult for an idiot to take over _the whole thing_ and ruin it, though.
Your position as a user on any given Mastodon instance is worse than it was on pre-Musk Twitter, undoubtedly (post-Musk it just has a CEO ruling by fiat, which I would argue is a worse situation than you have on most Mastodon instances), but it's unclear that your position as a user of Mastodon writ large is.
A mechanism to backfill posts would definitely be a good addition.
Edit: for clarity, ActivityPub provides an outbox from which another server (or anything else that speaks HTTP) can pull posts. Adding a backfill feature to Mastodon would not require any changes to the protocol.
In some ways, and for some purposes, this is very likely a feature. A good amount of the fediverse wants small healthy communities, not Twitter but federated (i.e. everything going everywhere all the time).
I will absolutely agree that it's sub-par for almost anything happening now, which is a problem. But it's experiencing unprecedented growth - there will be growing pains, just like non-federated things do when they have unprecedented growth.
What can you do if you have issues with the moderation team at Twitter? Honest question-- I've been off Twitter for over a decade and hear about everything there second-hand.
This sword cuts both ways for the fediverse (allowing bad actors to just move to another instance, for example) but on the whole I find it allows for a more people-centric approach to online community.
Keep in mind that Torrents tend to be much larger than your average Tweet/Toot/Whatever and that the interconnects between the servers need to transport only those messages for which there are subscribers, something that could be optimized for (you really only need to transport each message once per sender if there are subscribers, not once per subscriber).
https://techcrunch.com/2022/11/21/tumblr-to-add-support-for-...
Big, heavy doubt. This Twitter debacle has generated ridiculous, crazy amounts of drama and emoting, the likes I haven't seen on the net in ... a while. This kind of emoting and drama stirring only works when the pot is large. A storm in a small teapot really isn't much. It reminds me of the high drama that used to happen on large IRC channels run by teens in the early '00s, where you'd wake up and some portion of the server got muted or banned for some completely opaque but intensely personal reason. (I was one of the teens so I'm not implying like I was much better than the rest of them lol.)
I'm in 20-something small Matrix rooms/spaces and Discord guilds and none of them are nearly this dramatic. The kind of folks that seek engagement on Twitter IMO are in it for the high drama that you can only get on a large social platform. The folks that stay on the Fediverse are looking for something entirely different. I'm hopeful that some folks try the Fediverse and realize they didn't need Twitter at all (not in some sense of Twitter anger, but more that it's good for folks who want to socialize in smaller spaces be aware that it's very possible to do that on the net right now.)
FWIW If you're trying to build a small community and afraid of scaling issues, I still think hopping onto Matrix or Discord (preferably in that order) is a better idea. Synapse might be heavyweight but it's still lighter than running Mastodon, and if your community doesn't want to federate, you can run the much lighter Dendrite (or even lighter Conduit which is a bit more raw.) There's a large ecosystem of clients on all the big platforms for Matrix. Discord being a centralized, closed system has all of these problems solved as well.
It certainly needs less features. I'm sure quite a bit of Twitter's infrastructure is dedicated to ads and analytics.
Fwiw, speak for.. yourself? /shrug
I pay for my instance. So do many. Because it's on Patreon and fully funded right now. Yea it's small, around 16k users, but we pay because we enjoy it. Also it's ridiculously cheap to contribute, which helps.
Thankfully, there already is work being done to implement this kind of functionality, giving instances the option to serve images over IPFS gateways. Further decentralizing Mastodon needs to be a very high priority for it to be sustainable longer term.
Services like Twitter, Insta, FB, the like are in many ways incredible. We've forgotten all about that, but let's explore...
They're pretty much always available. They scale endlessly without you noticing. They are fast. They have apps on every touch point. They preserve your content pretty much forever, and this content exponentially grows. They have armies of people as well as advanced AI to filter out the most horrible of human depravities, so that you don't get to see it.
And for all of this, you don't pay a single penny. Yes, you'll get a few ads. You can even block those. An incredible amount of value offered, for near-zero costs.
But think about it. We have networks in which we can connect to any other connected human being. Post near-unlimited amounts of content which is reliably preserved. We have well functioning-apps and reliable services, and relatively clean content. For the price of: 0$, and ads.
What I'm getting at is that all this stuff that we take for granted, isn't a given or free when you scale Mastodon. It costs real money to run instances which hampers growth and sustainability. There's a steep human costs to moderation.
Interesting how you think that the only costs could be financial.
Also, the presence of ads is only tangentially related to impact on privacy (yes, another huge cost).
1. The fediverse doesn't necessarily need to scale to 100 million or more people, and if it were to hypothetically hit a hard limit at 10 million, that would still be fine. Anecdotally, many people are reporting similar levels of engagement--and more positive engagement--with even a small percentage of their twitter audience.
2. Usenet failed largely because it was overrun with spam. While Fediverse moderation is distributed and therefore inefficient, it's still done, so spam doesn't seem likely to overrun anything. In theory, each server moderator moderates their own server, and any servers that don't end up blocked.
3. At least at this point in time, people seem to be willing to pay for their own servers. Many servers are running patreon accounts or similar, and most I'm aware of are running a surplus of funds. Of course, that might not always be true. Smaller server are very cheap to run.
4. I wouldn't think of torrents as an example, but rather the web itself. So so so many individual websites are out there being paid for out of pocket, and nobody seems to be worried about them all failing at once.
Would you say the same thing about e-mail, that it "would still be fine" if it only ever scaled to 10 million people?
The fediverse, it turns out, isn't primarily a messaging service.
So yes, if you define X as "a messaging service that intends to connect people across the world," then X must support as many people as possible. But if you don't define X that way, then not supporting 100 million people isn't such a big deal. Millions of people are already happily on fediverse servers despite the lack of any current or former presidents there.
People who want to follow whatever Obama wants to say can do so in any number of ways. The Fediverse might be one of those ways, or not, and it wouldn't really detract from the Fediverse if it weren't.
I mean, I see stuff from the most recent former president all the time, and he's on neither twitter nor the fediverse.
If Mastodon isn't "a messaging service that intends to connect people across the world", then I can't even imagine what it's supposed to be. Because that is exactly what it is.
And if it somehow hit a limit at 10 million total users, then it would be a complete failure. And to say that "it wouldn't really detract from the Fediverse" if you couldn't follow popular people because some limit were hit, you couldn't be more wrong. It would detract from it entirely and totally. It would defeat the entire purpose. And just because you don't want to follow popular people doesn't mean other people don't.
Precisely because the entire point of federation is that you're not stuck with separate forums, each on their own server, that don't talk to each other. The entire point is that they communicate in order to connect as many people across the world as users desire.
Again, how is this any different from if e-mail were capped at 10 million users total? Since you didn't answer the question.
Barack Obama is not social on Twitter. You can't talk to him. You can't have a conversation. Look at his Twitter feed: when's the last time he's replied to anyone? He issues mini press releases, that's all.
You can't email Barack Obama either, by the way. I mean, you can try, but good luck with that.
You talk about a messaging service, but it's entirely one-way messaging between famous people and non-famous people.
Mastodon users aren't looking to follow famous people and be passive recipients of press releases from famous people. They're looking to have conversations with each other. In this respect, Obama is irrelevant.
There are tons of people on Twitter with over 10 million followers who interact with Twitter a ton. They're very "social".
If a social network can't handle people who are popular, then that's a real problem. And remember, people's accounts can start unpopular, and then they gain millions of followers over years. So what do you do, boot them off for the sin of saying things lots of people want to listen to and interact with?
They certainly don't interact with 10 million people. That's impossible. They only interact with a tiny subset of their followers.
> If a social network can't handle people who are popular, then that's a real problem. And remember, people's accounts can start unpopular, and then they gain millions of followers over years.
I disagree. The most popular person I followed on Twitter had 300K followers and had been on Twitter since 2006. It's really not a big deal if there are no celebrities. The celebrities can post wherever else on the web they like. Don't need them. Let them continue on Twitter, whatever. Not a problem.
I've spent literally decades on the internet, on countless forums, never interacting with celebrities. It's fine. It's great!
This isn't about "celebrities" (a word I never used), it's about popular people all around, like authors, journalists, and so on.
And sure nobody is personally replying to all of their 10 million+ followers, but those 10 million+ followers do want to follow! And popular figures absolutely interact with and reply to tons of their followers quite frequently, that's the whole point.
Frankly, nobody cares if you "don't need" anybody popular. But to argue that a place like Mastodon should make it impossible for anybody to be popular, is ludicrous. If Mastodon wanted to stay tiny for people like you, separated into little popular-people-not-supported bubbles with a some hard max limit on followers, then why would it even have been built on federation in the first place...?
The idea that an author with only a few followers could have a best-selling book, and then at some point enthusiastic readers try to follow them only to get "error: max follow limit reached", would be a total and utter failure. A bug, not a feature.
I don't want to quibble over terms. Celebrities, popular people, whatever you want to call them.
> And popular figures absolutely interact with and reply to tons of their followers quite frequently, that's the whole point.
And popular people have plenty of places where they can do that. But why does everyplace have to be amenable to people with 10+ million followers?
> But to argue that a place like Mastodon should make it impossible for anybody to be popular, is ludicrous.
That's a straw man. It's simply a question of whether it's technically/financially feasible to scale to that level. And there's no reason why it has to be. I don't see why it needs to be designed around the needs of super popular people.
> The idea that an author with only a few followers could have a best-selling book, and then at some point enthusiastic readers try to follow them
It's possible for a person to outgrow a place. That doesn't mean there was anything wrong with the original place. Popular people have different needs than non-popular people, but that doesn't mean we have to cater to popular people as a prime directive.
I find it ludicrous that every social space on the web would have to make itself accessible to celebrities (my word), as if it were a legal requirement like making it accessible to the disabled. Celebrities are not a protected class. ;-)
The sites that cater to celebrities are supported by advertising. They need all the eyeballs brought by celebrities. Mastodon does not have advertising. It doesn't have the same incentives. That's not a flaw, it's just a different set of choices and priorities.
This is where you're wrong. Twitter is the only place they can do that right now, where the audience is for that type of content. And people are positioning Mastodon as a better alternative to Twitter. So Mastodon needs to be amenable to people with 10+ million followers. That's just logic. Otherwise, people wouldn't be actively encouraging Twitter users to switch to Mastodon the way they are.
If you want a product that won't support popular accounts then that's fine, but that's clearly something different from Mastodon and the Fediverse then. Really, that's just what regular forums are for, and always have been. But Mastodon is not a forum.
Facebook? Instagram?
> And people are positioning Mastodon as a better alternative to Twitter.
People are positioning Mastodon as an alternative to Twitter that is not and cannot be controlled by Elon Musk. He is the reason people are leaving. So the last thing that Mastodon needs is to be amenable to people like Elon Musk.
You asked about email, I'll counter with actual message services. For reasons clear only to the people involved, I have contacts I can only reach via Signal, or via WhatsApp, or via Telegram, or via WeChat, or via Discord, or via Slack, and so on. How do I cope with this? I just do. Similarly, if I had to visit eight different web pages to track my favorite super-popular people, I just would. Or I'd do what I actually do, and follow almost none of them, instead letting people who are terminally online quote and link to the ones worth knowing about.
If ActivityPub filled up, and couldn't take on any more users, then I would first expect technical solutions like a sharp rise in relay servers, which combine the output of many servers into one outbox to be polled. After that, assuming there was some sort of hard limit that couldn't be overcome that way, I would expect that I would need to create a new account on some other ActivityPub-like network, or even a separate unfederated network also running ActivityPub. My chosen client (Mastonaut) can already support multiple ActivityPub accounts on different servers, so my experience would likely remain the same.
You initially suggested that the Fediverse can't scale. It's worth noting that Twitter itself fell apart long before it reached anything like today's mastodon numbers, and the "Fail Whale" was a constant companion in 2008, when Twitter had only one million users, 200k of which were active per week[0]. So already ActivityPub is scaling better across the network than Twitter itself once did all by itself. Maybe Mastodon (Ruby on Rails) will fade over time, replaced by something written in a more efficient language (like Twitter was). Maybe ActivityPub itself will be converted to use thrift or grpc or similar. Maybe ActivityPub can support 100 million users just fine. It seems to have handled going from 300K monthly active users to 2.5 million monthly active users without trouble[1].
The total Fediverse seems to have more than 8 million total users now. I guess we might find out.
0. https://techcrunch.com/2008/04/29/end-of-speculation-the-rea...
1. https://blog.joinmastodon.org/2022/12/twitter-suspends-masto...
people are doing things like this too. you hear about Mastodon and ActivityPub because it's the easiest to setup, has the widest network, and works "well enough" for the userbase that's grown on it organically over the last decade.
if you're passionate about a particular experimental/p2p design, go seek it out. Secure Scuttlebutt might be up your alley.
What fascinates me, is that LetsEncrypt has 2 beefy servers and that's it.
Surely, a few million packets of highly compressible text over the pipe isn't that much of an issue? Is it? I've never done it so interested in hearing thoughts.
Agree. I've been using Kagi for a few months now and have been pleasantly surprised I haven't had to revert back to Google once. I resisted for a long time, but Googles results have been going downhill for a whle, and I realised that if I want to use something, somebody has to pay for it, and that somebody should probably be me, the user.
"Unicorn" companies like social media live off investment dollars then usually ads. The former has dried up, only leaving the latter. I'd never pay $8 a month for Twitter, but I also don't care about it. I'd imagine many daily users would.
I've been a part of many communities where the owners have to beg for money for running costs monthly. It's always a struggle, and I don't see Mastodon being any different without compromising their principles.
While I love writing Ruby (what Mastodon seems to be mostly is build with) it's syntax and cleanliness, it always had an issue with scaling.
Some of the new users will have to run new servers.
For example if public officials are moving to Mastodon, I expect the state to run their own servers. It was never a good idea to allow Twitter to be a platform for (semi-)official communication.
If the government does not want to put public communication into the hands of tech billionaires again, their second best option after using available software is to build a platform from scratch.
Why doesn't anyone remember how shitty Twitter was when it was growing. Slow and constantly breaking.
Mastodon is working fine even after millions moved to it. Yes, Admins had to scale up, but they did it.
I like that I pay my admin. No ads. No bullshit.
I don’t see any normal person self-hosting their own email service. Even if they did, good luck with handling spam. It can only viably scale to hundreds of millions with centralization; hence Gmail, Outlook, etc to reduce / limit the spam issue.
Mastodon has the same mistakes as the other federated alternatives and has already proven to be unable to scale more efficiently than Twitter, let alone the ridiculous hurdles to begin using it.
We are talking about scaling to hundreds of millions, not ten thousand or a shy hundred thousand and it falls over on a single moderator’s instance.
But email didn't fail as a result. Just as you said, email services popped up.
Just give it a little time.
With centralization; hence the spyware, bans and spam prevention which only benefits the big centralized email services like Gmail, Outlook, etc.
Email was meant to be used as a decentralized protocol such that one can operate their own email service; but despite many here claiming to run their own, they gave up and many still won't. Even sending an email to a Gmail and Outlook addresses will just fail their spam filter test.
So it has been a failure for email without centralizated providers.
> Just give it a little time.
A 6 year federated experiment is not early days to see that a worse version of Twitter has already failed.
A really popular person might have many millions of followers on tens or hundreds of thousands of nodes.
That person then boosts a post from another server and they have effectively DDOSed that server. All the other nodes will start querying that server about that post or image.
You can say that a single user should not have that many followers, however the root of the issue is that mastodon will have some scaling issues.
They've had years to get this right. This is their moment to shine and they're blowing it.
So if the likes Obama or Stephen King want to move over to Mastodon their barrier of entry would be to complete this laundry list [1] or find a service that does this for them? Given that they own the instance, they would also need to provide moderation for it.
I don't disagree that there are solutions to the scale problem. What I have trouble seeing are solutions that are viable. That's the thing about a centralized service--you have a company that handles all of this so your users don't have to.
[1] https://docs.joinmastodon.org/user/run-your-own/#so-you-want...
Assuming that Mastodon growth continues, such services will probably emerge (there are already a couple of Mastodon hosting services, but there's probably a market for a "so you're a celebrity who wants a single-user Mastodon instance" service which has yet to be filled).
> Given that they own the instance, they would also need to provide moderation for it.
If it was a single-user instance, that would be fairly trivial :)
When you start off an an empty instance, the only thing you see is yourself. There are no trends, there are no posts, and everywhere you click is just you.
You have to start following people for it start populating, but that requires knowing strings to follow. And then you can you see is stuff in your bubble (maybe a feature?).
But it really missing the dopamine response, as after you read the stuff people you follow posted or boosted; there is nothing.
A dedicated person can push through this initial period. But for mass appeal, people will quickly bounce if they feel like no one else is there.
But that's fixable in several ways - local media cache for an instance, the boost including the preview rather than each client fetching it for themselves, etc.
The problem* with ActivityPub is that it's been designed and developed in a bit of user vacuum and it definitely has issues but they can all be worked on. Much like Twitter going from Rails to Scala, etc., to fix their issues.
The instances cache pretty aggressively, but you still have a thundering herd when a popular person boosts a message.
I suppose I should have been clearer - apologies. I was talking about a cache for previews (which would presumably be folded into the local media cache since it's already there.)
> you still have a thundering herd when a popular person boosts a message
Not if the boost includes the preview - that -should- cut it down significantly (and indeed it has been proposed but is still being bikeshedded by the Mastodon devs because "what if someone sends a fake preview?!?!?!?!")
What will happen is the same thing that happened to web servers and email servers. Businesses will pop up to run your instance for you.
With respect to monetisation, I hope the fediverse will provide a live sandbox to try all kinds of experiments at eye level: Free, ad-supported, fees, one-time fees, pay-per-toot, pay-per-followed-instance, whatever. Multiply by private, non-profit, and for-profit approaches.
You know how we use websites right now? Who's responsible for planning, scaling, and paying for all of that? Users have no interest in running their own "website instance". Right?
And we're already at the point where for big groups running your own requires multiple full time jobs.
Even if things don't work smoothly, it's fine. Most breakage is instance local and temporary. I've been on Twitter with a weekly failwhale and it's been fine too.
Each server in the fediverse should stay relatively small.
Consider running your own for your friends.
Even if you are still subject to the whims of your hosting company, you still retain a fair amount of autonomy unless you are running a community like Kiwifarms (in which case I find it hard to be sympathetic anyway), or your hosting company happens to be located in a jurisdiction with enforced censorship laws.
One of the benefits of the fediverse is that you can experiment with new payment models.
You're providing a real service to your friends - a mastodon server. It takes real time, real skill and real money (buying hardware or renting a vps) to do this. Your friends should compensate you and it's fair to ask them to do so.
How you do this is up to you.
I host a gameserver for my friends. This is a really nice bit of hardware in a datacenter. It costs me $600 a year directly and maybe 4-6 hours a month. I have about 8 folks who play on it with regularity.
For my friend group, we "compensate" via trading food, either cooking for each other or catching dinner out. That works well for my friend group and I perceive this as fair.
Your friend group may be different. You may literally ask them to throw in $5 a month or maybe you trade services (your barber friend gives you a free haircut on occasion, whatever).
There isn't a magic equation of what is fair or right for your circumstance.. and it's ok to experiment. Part of why the fediverse is good is that these experiments can be run.
The problem presented wasn't how you pay, it's the fact that you pay and take on moderation duties and curation.
Getting reimbursed by friends doesn't change that, and if anything it convolutes the latter part of the problem by leaving the door open to some awkward dynamics.
A centralized group chat that needs funding to stay alive sounds like a pretty awful product.
So it goes.
The point is that people went from an experience where they didn't take money out of their pocket every month unless they wanted to, and didn't have to curate for others or moderate for others.
I pay for a game server and won't accept a dime from my friends for it not so I can feel like king of some fiefdom, but so we can enjoy a game amongst ourselves.
If moderation is the part of that experience that you're cherishing, it's no wonder you can't relate to the complaints others have.
That was basically a vivid hallucination. You offshored the mod work to underpaid contractors in foreign offices and paid by selling your focus and privacy. You didn't have to feel guilty because some nice folks in fancy clothes with a big PR department told you it was cool.
Those folks in the fancy suits are realizing it's too hard to make money that way.
You can still chase that bargain. You will see more and more ads. Your privacy will only get more invaded. Eventually you will have to pay.
Big forums had big hosting costs and required donations from members, had issues with spam, had to source volunteer moderators from the community, etc., just like big Mastodon instances today.
Small forums had few of those issues because the scale was smaller. The problem was not that nobody wanted to run a small forum: it's actually kind of fun to be a forum admin for your friends, and only one person in the group actually needs to be the admin. The problem was that nobody wanted to be on a small forum because it was small.
Now you can have it both ways: a small instance with trivial running cost and moderation demands, but all the benefits of being on a huge forum with thousands of concurrent users.
See also: Discord servers. Administering a Mastodon instance that you pay someone to host is not much more difficult than administering a Discord server. If people found it difficult to moderate a Discord server, why would so many friend groups, gamer groups, content creators, sub-Reddits, and software projects have their own small Discord servers? The answer is that the cost of moderating a small group of people is really small because small groups are easy to moderate, and the main downside of a small group isn't really relevant when the small group can follow and interact with anyone in the huge federated network.
And yet many of them died or, better put, most of the new forums were never born nowadays because communities moved to centralized, managed solutions (Facebook, Telegram, Discord, Slack etc)
Apparently Tumblr intends to implement ActivityPub, so it might be happening sooner rather than later.
> Each server in the fediverse should stay relatively small.
Certainly doesn't vibe with the project owners pushing a set of "blessed" servers to every visitor: https://joinmastodon.org/servers
I think this is the biggest reason Mastodon will stay a fringe platform. No one who enjoys it will accept anything of its issues as valid criticism, and if you point out the fact that such a mentality will keep it from getting popular, the reply will be "good! I don't want it to get big!"
Meanwhile the project itself very clearly wants to get popular, and it's clearly being sponsored with the intent that it should become more and more relevant, much to the chagrin of those who were using it before it was cool.
It seems like eventually the cool kids who start to shut out the new kids to stave off an Eternal September, leaving the project in an awkward spot and a steady decline for everyone involved as the old guard shifts energy to building taller and taller fences around their respective echo chambers. (Now what was that about registrations again?)
Then supply and demand will dictate that people will start paying for accounts, and that will be a good thing.
Tumblr will be happy to host them! Matt[1] from Automattic has stated that Tumblr will be joining the Fediverse soon.