64
Fediverse sustainability
(lemmy.ml)
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 17 Oct 2023
64 points (78.6% liked)
Fediverse
28537 readers
507 users here now
A community to talk about the Fediverse and all it's related services using ActivityPub (Mastodon, Lemmy, KBin, etc).
If you wanted to get help with moderating your own community then head over to !moderators@lemmy.world!
Rules
- Posts must be on topic.
- Be respectful of others.
- Cite the sources used for graphs and other statistics.
- Follow the general Lemmy.world rules.
Learn more at these websites: Join The Fediverse Wiki, Fediverse.info, Wikipedia Page, The Federation Info (Stats), FediDB (Stats), Sub Rehab (Reddit Migration), Search Lemmy
founded 2 years ago
MODERATORS
That's like post #10 I see from random users proposing we should somehow run ads or whatever to finance big instances.
I haven't seen a single statement going in that direction from big instances themselves. None of those posts referred to anything.
Is it just overconcerned people worrying about things which are not their problem? I assume people who can run a big instance would notice if they are getting into financial troubles. As long as they don't speak up, I would conclude we don't have to worry. The current model (whatever it is) seems to work well enough. Did they ask for advice, do they need advice?
Maybe it's that people are so used to being forced to see ads and pay half their wage for insulin that they cannot imagine nice things exist.
I think we should try to keep it nice, and not revert to capitalist enshittification prematurely, without any necessity.
We currently have more than 1000 instances on Lemmy. Maybe some do run ads, who knows. You can join them if you like, or host your own.
Show the problem exists which you try to solve. Point to instances who struggle financially, who consider running ads, something like that.
See my other comment examining where the top 10 instances by userbase get their funding from and how well they're doing
Not to mention that over the years there have been a lot of instances that have gotten into a variety of precarious situations that could have been avoided or alleviated if they had a lot more money.
Their problem is that they allowed themselves to become too big and unsustainable in the long run.
It's not just their problem. Even if every instance carefully load-balanced users with each other so that all instance were the same size and nobody was too big, there would still be a problem securing funding as the fediverse as a whole gets bigger.
Donations alone on the biggest instances aren't enough to keep the lights on, spreading out those users across other instances won't make more money suddenly materialize, in fact it might make money disappear faster, as smaller instances have a higher cost-per-user due to insufficient economies of scale.
If the instance becomes so big, that it depends on donations, then it is too big.
Exactly, I'm surprised how little I've seen this pointed out in this thread. There's essentially zero reason for instances to grow beyond basic sustainability.
@petunia @Spzi Some are not about money: mastodon.lol is purely a personal decision; switter.at is not gonna lobby against governments that want to censor queer voices (which is what "online safety regulations" are really about). For Pawoo, Pixiv certainly had the money to keep it running, so this might be profitability concerns (given that at that time Pixiv also phased out other less-popular services to focus on its main platforms); CrossGate/Russell could be financial and liability concerns.
There are no "big instances" in the Threadiverse. The largest one (LW) has less than 13k active MAU. These numbers are ridiculously low and offer no real stress to the system. Let's 10x this number and see what starts happening.
The top 10 instances account for 74% of MAU. And the bigger instances (LW, Beehaw) are balkanizing the Fediverse: trigger-happy with the defederation buttons, avoiding any instance that can bring "unwanted" activity, etc. Even if other instances start making experiments, they will only be interesting if they happen out in the open.
Does it? From my perspective, we have a small group of people who are just messing around with things that they can run themselves, a slightly larger group of people who are discontent with reddit and wanted an alternative, but very few people who actually care about an alternative and are willing to put substantial resources to help with development and to accelerate adoption.
People are not forced to see ads. Ad blockers exist. Which in a way is actually a problem. People managed to enjoy sites and blocking ads, so they got used to the idea that no ads + free access is an universal possibility and the natural state of social networks.
As for price: I've been offering plans for Mastodon access that cost $0.50/month/user on communick. I've had exactly ZERO people on this plan.
My conclusion: it's not the price point. It's just that people don't want to pay for social media.
Have you considered that it isn't the price but the subscription? Many people I know have a real aversion against subscription services with this constant threat of being cut off and arbitrary price increases.
I am pretty sure an up-front single "life-time" price would have more takers, even if such a promise is obviously still subject to many caveats.
Subscriptions have gotten a bad rep lately because of companies that try to turn a product (like a car and heated seats) into a "service", but there is nothing inherently wrong/unfair about someone that provides a service that has occurring and constant costs based on usage.
Also, for those that want to have full control over their own identity, they can have their own managed server, which is still a bit expensive but will be made a lot cheaper with the next generation of fediverse services (like takahe and mitra) . Once that gets more mature, users would be able to bring their own domain and a service provider would be a commodity like an email hosting service. In the case where I can port account and my identity to different providers by simply changing a DNS record, the power will be fully in the hands of the people and there will be nothing for them to worry about.
Yes, because even with federation it is inherently advantageous for a user of a social platform to be among the largest pool of people they can identify, to make random stumbling into discussions and groups as likely as possible.
It's a weird thing where we want the federation to provide a network of smallest scale platforms, yet we do this for social media, where the experience is naturally best when it starts with a single giant platform you filter down not an ocean of individual bits you have to glue together.
I'm guessing you are not old enough to remember how the internet used to work before Facebook? This idea that walled gardens are somewhat better is a meme that needs to die.
The web itself is the giant platform. Ease of discovery is not an inherent property of centralized networks, it's just that we haven't had built the proper tools to make this work in a decentralized manner.
To make my case: what killed RSS was not that it was difficult to discover new blogs. What killed RSS was that it couldn't be monetized by the publishers when they started using it. What made Twitter so successful was that it let those publishers to have some sense of control over the distribution. Had we properly supported content creators with actual money instead of the promise of eyeballs, the internet would be a much better and healthier place than it is today.
Geezus. I'm old enough to remember how the Internet was before the Internet. Sorry. 😅
You however completely missed the point. Social media in its nature benefits from centralized approaches. As a use case. Independent of who operates it. Users have it easier the more central it is. It doesn't need to be walled. Of course not. But it should be centralized.
Again, I really don't see why. Content/Peer discovery can all be made transparent for the user, addressing and distribution as well.
I see the benefit for those building the platforms which in turn make these networks more attractive to users, and I see how the overall cost of the whole system is lower if it is centralized (economies of scale and avoid redundancies), but I'm failing to see how (all else being equal) the users benefit from a centralized system over a distributed one.
The fact that you are on LW and I'm not does not stop us from doing anything on Lemmy that is possibly only on reddit. Why benefit would there be for me to join LW?
Like I said above, specifically for the "I want to socialize" use case of social media sites, there's no upside to federation. It makes discovery harder, and a giant portion of what made Reddit so amazing was the random stumbling into things.
And yes, sure, federated systems can be made to more closely emulate such a centralized approach, but that's why I said it that way: A centralized pool of social media content (for a given social media platform) is beneficial to the user, they can randomly stumble into topics and groups, and filter things down to what they desire.
In an ideal federated system, that is in turn exactly how the content would look for the user: They'd not even realize the content isn't all on whatever instance they're on, it's fully transparent. Because that's easier for the user. No matter how low the barrier to finding federated content is, there's still no upside for the user having to take that step and go hunt for federated concept. From the perspective of the user, that is.
It's not a big issue of course, but it does mean that by default, more users flock to where there are already more users.
I think we are talking past each other. I understand that the current implementation of the federated social media lacks a lot of things, so I am not disagreeing that currently people would benefit from joining a bigger instance.
But my argument though is that we can have federated social media does not tend to overcentralization. If the Fediverse gets popular - really popular, tens of millions of active users popular - then there will be too many independent actors who will be participating and the whole system will lend itself to many different hosting schemes:
IOW, if things grow and it becomes a viable alternative to the status quo, it will end up as a core infrastructure component, like email. And yes, Gmail is by far the largest provider and the hold a lot of power, but even they can't simply decree to flip the tables in their favor.
The hell there is.
When "I want to socialize", I don't have to go march up at the UN headquarters, the Vatican or whoever "controls people" and ask for consideration. I can literally just walk out my door and walk to my local plaza. Or maybe the local grocery market. If I'm feeling lucky I can take the first googelbus and go to the nearest stadium, anime store or vintage disco bar. None of those needs to rely on the fact that other supermarkets or stadiums exist to provide socialization.
I disagree... as a user you want relative proximity, but not centralisation. Which is exactly what federation provides.
Think of it like stores:
The centralized social-media is a bit like those big-box stores that have a little of everything. Hard to navigate and find the stuff, usually they only offer the items with broader appeal, and the entire experience is just unpleasant.
The Fediverse is more like a mall with many smaller shops. Small niche shops can survive because the many other shops drive foot traffic and if you are not interested in tools for example, you can just not enter the hardware store.
I agree with your overall sentiment but am I crazy or is RSS more popular than ever?
I doubt it is overall, but I've certainly seen more talk about it lately than I ever have. Not surprising considering how many reddit refugees I've been chatting with.
Is it? Maybe in absolute numbers it has gone up, but I remember when established newspapers and journalists would write on their blog and have full-text feeds, while nowadays everything seems to be on substack/medium and the RSS feeds just puts out a link to the gated content.
If you include podcasts, which are delivered via RSS by definition, undoubtedly RSS is more popular than ever.
It's a little disingenuous to do that though, so in this context we probably shouldn't count it.
Point where I said we should run ads
I didn't care so much about specific wordings but answered to the gist of it. Yes, I cannot strictly quote you on that, but so what?
"Many are opposed to ads" gave the impression it would be worth considering to have ads.
Anyways, that's like the least interesting angle, to discuss what specific words you used.
Before we revert to ads surly we try medals. Set a standard price on each. Then when a comment/post recieves such a thing divide that reward between user, user instance, post user, post instance, community, and community instace. That way servers, admins, and high quality content creators all have an incentive. It could theoreticaly be weigted however wanted. Only issue I see is it would need some sot of blockchain to ensure no fuckery goes on.