This is the issue with many of these types of services. Geeks love them; but no one else does. I don't like running a VPS or an AWS instance, because of all the overhead, so I tend to use shared hosting. I'm quite capable of running a VPS, but I can't be arsed.
I suspect that we'll be seeing some fairly large corporate Mastodon servers (like...say...Twitter). These are likely to be popular.
Once it was obvious that emailing was a thing many people wanted to do, and a free server became available, servers started springing up all over the place, many more implementations of the underlying protocols were developed, businesses figured out various ways to monetize, the standards evolved (somewhat).
Now let's do an inventory on ActivityPub...
1) It became obvious that tweeting was a thing many people wanted to do? Check
2) A free server became available? Check
3) Servers started springing up all over the place? Check https://www.similarweb.com/blog/wp-content/uploads/2022/12/m...
4) More implementations were developed? That's where we are now
5, 6) Stay tuned
- With Mastodon, it's immediately more complicated than that.
- Free isn't a factor in Mastodon's favour because Twitter, Facebook, and Instagram are also free. And it's simple to understand and join.
- Everyone is on Twitter, so if you want to see their tweets, you have to join too.
Is there something inherit in Mastodon that makes it more complicated, or is this an obstacle that can be overcome (like email did)
I'm writing a very basic social media app, myself, right now, and have developed an appreciation for the complexity behind the façade.
This is why we don't see normal users (non techies) themselves self-hosting their own email servers to send a email, even though the option is there.
Even if they did, the biggest email providers deal with the spam problem better than self-hosted individuals, hence why it isn't a problem for users signing up to Gmail.
Mastodon will naturally gear towards centralization once the larger instances open up their instances and add defaults, ruining the whole point of federation with the biggest instances being at risk of buyouts and takeovers offers.
What I'm referring to is free on the supply side. There is now free and widely available server software which not only will let you host a social network, but will also connect you to an existing social graph via ActivityPub. Prior to this, how were you going to compete with Twitter? You needed to build your own client and your own social graph and that was a huge barrier to entry. The tools which exist for a new competitor today (e.g. Mastodon and still a fairly small fediverse) are far from perfect, but they are improving and you can extend them however you like. That looks a lot like an environment where someone disruptive could appear, add value is missing, figure out a way to monetize and so on.
Not anymore and less each day. Twitter is no longer a valid choice for a lot of people.
You missed some steps in your inventory listing though:
0)Able to communicate with anyone else with an account, regardless of which server they have signed up with? [ Nope. ]
Also, like BTC, the whole thing is network effects, and barriers to entry dramatically curtail networks. People like Twitter because celebrities and journalists use it, not because of the data model.
That's what ActivityPub integration is going to look like, IMO. With companies like Flipboard and Tumblr adding integration, it's undeniable that the network is growing (even if it's not it's strong suit). For people who want an RSS-style feed with light interactive elements, it's a dream come true. Maybe that encompasses Twitter users, maybe it doesn't.
Aren't the most popular ones already ran by companies?
As far as I can see from the data, the most popular is mastodon.social (ran by Mastodon gGmbH), followed by Pawoo (Russell Co., Ltd / formerly Pixiv Co., Ltd), mstdn.jp (Sujitech, LLC), baraag.net (independent) and mastodon.cloud (Sujitech, LLC).