Cohost to shut down at end of 2024
cohost.org
cohost.org
It started on a thread asking how much it costs to run an instance per user. Because most admins runs things as a hobby, all answers were like "oh, my instance has a few hundred users and I run it on $20/month VPS, so it's just a few cents per user per month"
No one even considered factoring the cost of the developers, the hours put in by admin, the value of the work done by moderators. Nothing. The only thing that has an actual price is the stupid VPS, so this is all that they think should be chipping in for.
For completely unknown reasons, there is a big overlap between the people that make this type of argument and the people that are mystified about the fact so many smart people end up using their talents to optimize the amount of ads being pushed to people online.
This is why I make it a point to pay for software that I find useful/enjoyable.
One of the most obnoxious side-effects of the VC startup mill's use of "free" services to fuel growth (loss leaders) is that they tricked people into thinking that software was cheap or free to produce.
Now people feel entitled to your efforts for as little as possible and, in some cases, get absolutely indignant that you'd ever try to run a real business and profit from your efforts.
Premature over-engineering, or optimization costs a lot more developers, administration and hardware.
At times it reminds me of when someone is using AI to do sometimes a SQL query might be able to.
So much effort avoiding SQL, to the point of using non-relational databases... only to relate them.
Everyone here has hosted something and has a take on how to golf those costs, but comparatively few have ever made payroll. So that's the discussion we get.
exercism is not shutting down. He responded to multiple comments saying he wasn't and he doesn't understand how people jumped to that conclusion.
OT: Don't get me on my diatribe about reducing drug costs by ending direct to consumer advertising.
Say what? Are these places known for cheap labor or is this a snide comment?
Now, imagine if these sites worked in a system where everyone would have to pay a little bit (or have someone else paying for them) so that every user is contributing to development/hosting/etc. Get people paying $9/year ($0.75/month) and I can bet you that you'd be a lot more than 3.5 FTE, we would be getting a viable alternative to Reddit and without all the enshittification and Surveillance Capitalism associated with Big Tech.
You could say the same of almost all corporate social media, and I am phasing them all out too.
For me, lack of federation meant they were dead on arrival, just like every Google product can't be trusted to keep existing (and I would never start using a new Google product). Any individual fediverse project could shut down, but the fediverse continues and is probably unkillable like IRC or RSS at this point, even if it's not as widely used as one might like.
If you want a way to "access more users and content", you can already make an account on a Fediverse instance and access those users and content there. What does Cohost offer by being yet another instance with custom software?
If the vast majority of the content I'm accessing is from other instances/services, why would I pay Cohost money for the privilege of having them act as an intermediary?
One specific problem with federation/decentralization is that you've now decentralized moderation and given up control over your service's culture entirely. This has some upsides and some downsides, but it again puts you into a weird position: If your moderation and culture position are identical to Mastodon's, why do you need to exist? If you're just another Mastodon instance, why would anyone give you money?
When I post on HN or Twitter or Tumblr or Cohost, I at least have a good sense of how the service is moderated, what the rules are, etc. When I post on decentralized services, at any point I can discover that oh, this random instance defederated the one I was unlucky enough to sign up on, so half of my mutuals can't see my posts anymore. Don't worry, you can just migrate your account to a different instance with different rules, and hope IT doesn't get defederated! And because each instance has different rules and culture, you get to look forward to people from other instances complaining that you aren't complying with their rules. It's messy! It's not fun!
Of course the rules on those services I mentioned aren't necessarily going to suit everyone's tastes. But I think that's good - a social media service doesn't need to be For Everyone to be successful, and being For Nobody is a horrible outcome.
At the end of the day I see the appeal of defederation but it simply doesn't make sense as a way to spend your engineering dollars if your goal is to be profitable. It Doesn't Make You Money.
Narrator: It could not have helped, it would have mostly just made them even more burnt out, because they (as they said in the post) were already on call effectively 7 days a week and have been since the site started.
> the fediverse continues and is probably unkillable like IRC or RSS at this point
Who cares? When an IRC channel, or phpbb forum I use closes down, or an RSS feed I used to read goes dark -- I do not think about the wonderful IRC protocol and how ossified it is or about the sanctity of DNS. And these things happen regardless of what the underlying technology stack are. I think "It really sucks that my community and the people on it are gone now." A domain that doesn't resolve and an empty mastodon instance are effectively the same.
The users who use the platform are why the platform is valuable, not because it uses some dumb particular tech stack.
I am asking because I happen to run one of those (https://communick.com) since 2019.
Even now I can’t find your service in the first page of my search results. I ended up just setting up a recurring donation to Mastodon.social via Patreon.
Your service looks very cool. Just pointing out that it is very hard to discover even for people who are willing to pay, as the GP comment notes and hopefully providing a useful experience report.
To be honest though, I think that Mastodon is an evolutionary dead-end. It had a huge head start in the space, but I am reasonably sure that it is not the future of the open social web. I am more inclined to pivot into a multi-protocol client (like Pidgin) and offering ancillary services than trying to pick a champion and invest into promoting it.
Interesting, this was my layman's read too after a while of trying to get into it. To me it felt like it was opinionated but in ways that I disagreed with.
For example, instead of a separate UI for private messages, they just add another option to the visibility selector. From a technical perspective I suppose this is somewhat elegant, but from a UX perspective it felt wrong. I saw a thread where people disagreed with it and the maintainers told them to pound sand. Similarly with quote tweets.
I know you mentioned not trying to pick champions, but are there any alternatives that you thought were particularly interesting? I'm willing to put in a little time to try things out.
The other thing I want to note is I just tried to use the "Set up Auto-Pay" button on Communick and got the error "Could not set up auto pay. Please contact support". I'd love to sign up though.
But I would pay if I had to, and I am considering paying for a GoToSocial server to experiment with an allowlist network:
https://codeberg.org/oliphant/islands/src/branch/main/ion
My server: https://jawns.club
Our finances: https://opencollective.com/jawnsclub
We're currently paying for managed Mastodon hosting on https://masto.host/
Finally, I think pay what you want is better so long as it works, since it doesn't exclude people who don't have money, but do contribute to the community in other ways. The only real reason to move away from pay what you want is if it doesn't pay the bills, and our finances are fine for now.
I can see a place for mandatory payments if you're providing extra services at a steeper price, such as paid moderation, but I think the number of people willing to pay what moderation at a living wage actually costs... is rather small. Perhaps if we made moderation more efficient, e.g. sharing moderation decisions between servers, paid moderation could become more affordable by splitting the cost between more users, but there are several problems with that approach... one being that moderation by members of your community is always going to be more clueful than moderation from outside your community.
Follow up question: if everyone treats social media alternatives as a hobby, do you think that it has a chance of being a viable alternative to the Big Tech platforms?
I think the profit motive is literally destroying the world given the climate crisis, as well as destroying everything good or useful about the Internet. So I think figuring out how to do things without the profit motive is the only way forward, as impossible as that might seem. Surviving capitalism with that attitude may be challenging, but we don't have a future under capitalism anyway.
I think that if the value of social media comes from the network effects, then it's not a question of whether you can find some core of users who consider it more than a hobby. (Ugh, you want a network for professionals for whom it's their job?!) For social media to reach its maximum value, it has to reach literally everyone, and at that point you're talking about something that ought to be a government service, like the postal service. For smaller networks that are less focused on reach, maybe hobbyists are the ideal providers.
It's amazing how we can get two people to accidentally take the same action while fundamentally disagreeing on everything else.
Short version:
- Capitalism is not the problem. Corporations that can take their profits in one department and use their vast resources to drive loss leaders is the problem. Break down big corporations into smaller ones that need to compete, and I can bet that we wouldn't see pointless growth and needless consumption.
- "do you want a network for professionals for whom it's their job?" No, I want the service to be run by professionals who can make a living out of it. The people running the service need not to participate in it. Like email or phone service, you don't expect to be free and you don't expect to be talking with the service provider daily. It should be a simple utility.
- "Public funding" is magical thinking repeated by people who don't understand basic economics: specially if you are "against growth" (like your comment about the climate crisis seem to suggest) then where is the Government going to get the resources to pay for developers?
You don't provide a foundation for the counters to the arguments the GP makes (f.e. you propose smaller corporations wouldn't have "pointless growth and consumption", without substantiating how you feel such a thing is possible with capitalism as described and commonly understood).
You call "public funding" magical thinking, but in the point right before that you mention something should be "a simple utility". Those are publically funded.
And of course we should not forget these are strawmen and/or personal attacks. This is not a respectable post that adds to the discussion, it's one that distracts and makes everything categorically worse.
Even on places where these things are run by public companies, they are not public funded. It's not like Governments give these for free and and/or come out of any government budget.
I never really understood this—what is their anonymous and secret funder going to use the source code for Cohost for? They've been very closed lipped about that part. Oh well, I guess we'll find out soon.
I guess I just don't understand what problem Cohost solved that other solutions didn't solve better and cheaper.
I don't particularly like any of the current centralized options but I also don't particularly care about federation and as such don't want to deal with the added friction that it brings.
> just go to the cesspool of
Cohost didn't feel like a cesspool to me. That's a pretty darn strong differentiator.
There are other UX differentiators for me, but it's all subjective. Cohost didn't feel as "cramped" as the others. The others really put the focus on the "micro" in "microblogging", even if technically they (now) allow for longer-form posts. The culture just isn't there for it.
Was this due to something specific they did?
Or was this just like how small subreddits can/could be the best place on earth, but then soon after they reach around 10,000 subscribers the quality of the content on the subreddit quickly drops significantly.
Something like the Eternal September effect:
> The periodic flood of new users overwhelmed the existing culture for online forums and the ability to enforce existing norms.
1. There's no global firehose that your posts automatically get dropped into. I'm a generally shy person, and one of my particularities is that I don't want my content (or, well, existence) to be pushed at somebody who didn't specifically ask for it. The number of social media platforms that meet this criterion is vanishingly small, because everybody and their uncle wants an Algorithm (tm). Mastodon instances also have a default instance firehose; I'm told people don't really look at that, but the fact that it's there at all makes me overthink posting, and I end up not posting at all.
Instead of a global firehose, you can add free-form tags to your post that act like much, much more localized streams of posts. You don't have to use tags at all; and you can also use completely nonsense tags that are more like footnotes, in a very Tumblr style that I enjoy.
2. The UI is wider than Twitter, or Facebook, or Mastodon. It feels like it's more inviting to longer-form content, and while there's plenty of short blurbs and memes to be found, there's just a much higher proportion of blog-style content.
Also, the way reposting and commenting interact is just way better IMO. If you repost something, the comments on the original post are visible above the comments on your repost, and that makes the whole chain of interactions feel more cohesive. In contrast, quote-retweeting on Twitter creates entirely separate bubbles of discussion (often at odds with each other). And on Cohost, while you can create a repost chain to make a kind of "thread" equivalent, the UI naturally discourages it in favor of just making a single long-form post.
Oh and most importantly for me, it allows html/css in posts, which enables so much cool stuff and creativity generally reserved for personal websites.
Got a recent link for this? I wouldn't think that would work so well.
Given the recent issues with Twitter, my country's president started posting on an official account on both Bluesky and Threads. Knowing about that ActivityPub support from Threads, I decided to try following him from my Mastodon account, which is in a server which does not block Threads (it can follow the USA president's and vice-president's official Threads accounts just fine), but that didn't work. What I found out after a quick web search, was that this feature is only available for users from the USA, and even then, it has to be manually enabled on each Threads account.
Which works for me.
I've seen several similar pending-shutdown notices for other sites/services.
> Cohost (stylized in all lowercase letters as cohost or cohost!) is a social media website publicly launched in June 2022[1] that is currently in beta, and will be discontinued at the end of 2024[2].
And right you were!
> The universe was (and perhaps was not), is, will be for the foreseeable future (of which the timespan is indeterminate) and might become (with perhaps a period of not being) all of space and time and their contents.
Their costs were super reasonable as an ongoing concern too compared with my experience at various startups. Losses that small are practically pocket change. If they weren't able to win investors over with their service's relatively strong performance since launch, what hope do these other services have when the investor money dries up? Do they have secret sauce that isn't "sell customer data to OpenAI" and "run more ads"? The former works right now but its days seem numbered as a realistic revenue source, and advertising becomes a worse revenue source with every passing day.
i.e. from the sources I've seen, Bluesky is operating on something like $20m in total investment. That's a good amount of money, but what happens when it dries up? What's their revenue plan? Whatever it is, I hope it's more compelling than Twitter's answer for keeping the lights on, more successful than Cohost's model, and more substantial than the Fediverse's.
From what I can tell the nail in the coffin to Cohost was having their big revenue feature unceremoniously killed by Stripe, which is sobering. Always sucks to have to put your fate in the hands of payment processors, since they have a long track record of being capricious.
According to:
https://bsky.social/about/blog/7-05-2023-business-plan
It's $8M, and their first plan is to sell custom domains.
Server details: https://bsky.app/profile/why.bsky.team/post/3l3232j6m3d2b
And: https://bsky.app/profile/why.bsky.team/post/3l322qwltzk2u
>> "we can probably deal with close to 10x from here, things may start melting at that point though."
Looking around the website I couldn’t help but notice this attitude: “we hate the software industry and we hate social media”… so they decided to start a social media company? I’m curious what their vision of success looked like.
To me, the fact that they burned out feels like the logical conclusion from where they started from.
They put their money (and investors' money) where their mouth was and actually ran a co-op instead of relying heavily on freelancers or cheap outsourced labor. It's unfortunate that doing that hampers sustainability, though.
That all being said, I don't think this is a huge indictment of Cohost—they had a solid 4 years, and in all likelihood they'd have had to close up sooner if their personal financial situations were more precarious. But a situation where making less than $94k is "precarious" is a very tough one to be trying to starta new social media company in.
It turns out that when you actually pay money to people for their work, it is expensive. That is because the value provided by the labor of humans is, in fact, valuable. However, if you instead just exploit people and take advantage of them, lots of things can be really, really cheap. In fact, they can be so cheap it'd be insane not to take advantage of, some might say...
Anyway this is what happens to a community that builds itself on closed source centralized platforms. When the maintainers decide to give up the users and community have no say.
Storage, as an example. It balloons. With video even more. Ever growing costs not necessarily on par with income. Hate FB as much as you like, but my photo posted 16 years ago is still there.
Depending on the software, social media can be compute intense. Not cheap.
Compliance. The legal headache of complying with every jurisdiction
And of course the worst: moderation. Deleting clearly illegal content, gore, scammers, spammers, impersonators followed by the large gray zone realm of the culture wars. A 24/7 operation that you can't ever walk away from and hollows you out over time.
All of that for what? A little carved out box for "your" people? Just open a normal forum, Reddit, Discord and piggy back on another platform.
You're carrying a disproportional burden for questionable benefits. You're never going to rebuild social media at scale from scratch. You're effectively creating a small chat group for like-minded individuals.
Mastodon is the biggest attempt at alternative social media. It's 8 years old and has 800K MAU and that MAU is perpetually tanking. It carved out a little corner for misfits. Good for them, but the entire "fediverse" is unlikely to account for more than that.
Regardless, too much expense for too little revenue.
So, if your idea of "can work" is "it works for me and my niche group who all treat it as a hobby" it is already working.
But if your idea of "can work" is "gets us rid of all social network that makes money by selling eyeballs", then the only way to work will be though finding a valid alternative business model.
That either means "moderates so that everyone posting agrees with what I mean by ethics to a satisfying degree" (and folks do not agree on this)
or just "I have fewer complaints about the moderation of this social network than the other popular ones".
To boil this down to what I actually mean with an example: What are the "ethical" allowed things to post about climate change on your social network? (please don't actually fall into the trap of actually discussing climate change here). I can think of about three or four different answers to this question that large blocks of people would passionately defend.
And so either "ethical" social networks don't exist or there is only one true ethical way to do things and that way is people have to agree with my position and lots of people who don't agree with me are very wrong about ethics.
You can do this for a lot of topics and you'll find that contrary to US politics, there are not actually only two ways of thinking about the world and one of them awful.
That is absurd.
For me, primarily one does not use a business model that profits from outrage, clickbait, misinformation, etc. Business models that use view-based ad revenue are unethical, for example.
Beyond that, it should make an effort to stifle the spread of harmful content. Where that line is drawn is of course hard to define. Twitter and Facebook largely ignore their responsibilities here, which is one reason I don't use them.
But this really is a bummer -- they had admirable goals, and I know lots of people who ended up there and who enjoyed it a lot.
Kinda like a chat app with two people, nice to use occasionally, but it doesn't get nearly enough of my time to be of much value.
Things like that have to be pulled into some aggregator (even if just by federation) before they reach critical mass, I guess.
-Emily