The difference I’ve noticed is that average people memorize how to accomplish their tasks with software and savvy people look for something that will accomplish their task.

1. People install communication apps because someone they want to communicate asks/tells them to. I don’t want to use Slack, but people who pay my bills use Slack, so I use Slack. I use Matrix too because I believe in it philosophically, but the UX is a little rough and very few conversations I want to have are actually taking place there.
2. I don’t really understand this one. Adblockers are relatively easy to install and the modern web sucks without them.
3. I haven’t encountered much of this in a long time, but familiar feels safe to people. Something they haven’t heard of might feel sketchy.
4. This sounds like someone trying to persuade you to install an app because you using it would be convenient for them.

It’s not unique to Xitter. Facebook has been deprioritizing posts with links in them for a while.

A person is smart. People, not so much.

I’ve been using Maddy for about a year. It’s easy to set up and has been trouble free.

Here’s an explanation from the Associated Press. The penalty is usually a fine, which impacts stockholders by making the stock less valuable and could lead them to remove board members or demand the termination of executives. It’s rarely used, but there is a corporate death penalty.

With a couple niche exceptions, AI hasn’t started making money. What it has done is attract venture capital investment.

Venture capitalists are driven by a fear of missing out on the next big thing. A billion dollar score pays for a thousand bad million dollar bets, and AI that lives up to the hype could be worth trillions. This is also why every existing tech company is scrambling to add an AI thing to its products even where it makes no sense.

I have seen that lemmy.world apparently has a very poor reputation among other instances.

It’s the largest by far, with five times the monthly active population of #2. One of the main things people want out of federated systems is decentralization, and having one huge dominant server goes against that goal.

I should note .world wasn’t the biggest when I signed up. I picked it because mastodon.world was a known quantity, which led me to believe the same team would run a stable server.

Most servers don’t micromanage community moderation. That sort of thing could happen almost anywhere.

When I say “by default”, I meant the vanilla Mastodon web client. Of course alternate clients could do just about anything.

It’s quite resistant to any single entity’s censorship, but if you share things most server admins consider unacceptable, other servers will block your server.

lemmy.ml… copy a sentence from the link provided which links to some article called “The Principles of Communism”

At least one of the Lemmy developers is a hardcore communist, and some people see lemmy.ml as a little sketchy for that reason. I see you found another server, which is exactly how federation is meant to work. While the overall culture tends to be left-leaning, most server admins are not hardcore communists and don’t censor political positions that aren’t advocating violence or discrimination.

I’ve self-hosted Pleroma, Hubzilla, Mastodon, and Wordpress with the ActivityPub plugin. I found Pleroma to be the easiest of those.

I think I’d recommend Akkoma today; it’s a fork of Pleroma.

I don’t think long posts in Mastodon are a bad thing at all. I self-host and I changed the character limit to 50000.

By default, Mastodon will collapse long posts in feeds. If you don’t want to see long posts, you don’t have to click to expand them.

If you have millions of people on a social network, and you go looking for toxic shit there, you will find it.

Well, on Mastodon, you might not because by default it doesn’t have a useful text search feature. If you’re on a server running a modified version, or something else with decent text search, you might. My self-hosted server was on a relay that briefly pulled in content from a famously toxic server. At first, I didn’t see it because I didn’t follow those accounts, but later, I added an improved search feature and tried searching for some terms of abuse. I did find a few absolutely vile posts.

Bluesky has had a working search from early on. Turning off some of the default moderation filters and searching for terms of abuse does, in fact find people using terms of abuse.

I’m puzzles as to why anyone would routinely post threads to Mastodon rather than moving to an instance without a short limit.

I’m not going to make a very good case for a position I consider terribly misguided, but in short it’s because trolls on other platforms search for terms related to people they want to harass.

Some people have also raised objections centered around privacy or consent to process data. I find that misguided too; ActivityPub is radically public in that it sends every action to a bunch of other peoples’ computers, and there’s no explicit consent to do any of the other stuff a server might do with posts.

Of course any server owner can just select * from statuses where text like '%search string%'.

Search and discovery are a big deal. It’s much less effort to find people and content, and to be found on Bluesky than on Mastodon/ActivityPub. It takes work on Mastodon, and I don’t think Pleroma, Misskey, and their forks are doing much better (though I believe Akkoma has a decent search function).

This is, in part because there is a very vocal contingent of ActivityPub users who do not want working search and discovery. I think that’s equivalent to saying they don’t want the network to gain mainstream popularity.

It does look like it’s possible to independently host all the components of the ATProto network now, so I’m hoping we’ll see a bunch of services pop up soon such that the network becomes more resistant to enshittification. I’ve seen Whitewind and Frontpage, but I don’t fully understand the federation model and haven’t been able to cause one to display posts from another.

It’s kind of not. It does appear to be at least theoretically possible to self-host any or all of the major components. Unlike ActivityPub projects, however, it doesn’t seem like anybody is doing that and offering services to the public.

I don’t think that’s necessarily true. What I do think is true is that there’s a chance some AI thing will be a trillion dollar investment, and the most motivating thing for VCs is fear of missing out on a giant score.

A nonprofit open source profit ought to have different motivations though.

That’s pretty much what I wrote in the comment box. The options for the multiple choice questions don’t really acknowledge that as a preference people might have.