yhvr

In case it helps you understand: barriers of entry to Mastodon, from my perspective as an ex-Twitter user and current Bluesky and Mastodon user (note it took me 4 tries over 5 years to actually "enjoy" Mastodon)

• Lack of content discoverability: Bluesky has programmable feeds and starter packs (although the latter is pretty flawed), Twitter has The Algorithm. When you first create a Mastodon account, you're not given much.
  • The primary reason Mastodon "stuck" with me on the 4th go around was this. I manually imported a bunch of my connections from Bluesky.
• Choosing the right server: on all my previous attempts, I was overwhelmed by server selection and went with generic all-purpose instances. Having a local timeline with things that might actually interest you is very helpful.
• The network effect: I don't think I really have to elaborate on this one

Bonus things that probably don't serve as barriers to entry but are real issues for active users:

• Lack of features: it's not possible to search for posts on ActivityPub at any reasonable scale. I don't know all my Mastodon lore and think I've heard that search was intentionally not made good in the past for privacy or something, but if you're posting something on the "blast my posts everywhere" protocol, well... it's still out there.
  • There's also no quote posts, which haven't been implemented for similar privacy/anti-harassment reasons. But instead, other software like Pleroma works around this by simply giving a link to the original post—the worst of both worlds, bad UX and not even an option to attempt to detach or prevent quotes.
• Fedi meta / sudden defederations: If you picked the wrong server—or any of your friends did—and you end up with a tempermental admin, you could unexpectedly lose contact with each other due to defederation, and you might not even notice.

This is not to say other platforms are perfect. There are plenty of things I don't like about Bluesky. This comment is specifically about Mastodon.

Distilled versions of Deepseek are available that can be ran on consumer-grade GPUs, and I have done this myself. I've even ran a really small one on my phone, though obviously at that scale it's going to be slow and bad lol

Replace "x.com" in the URL with "xcancel.com" to see context, replies, chronological user tweets, etc. It's a Nitter instance that still works.

Hadn't heard of pikvm before. Will keep that in mind, thanks!

While you didn't name names of what app you were using for streaming, I just got into a similar situation with my dorm and what I found worked was using wired ALVR for my streaming. Not wireless, but good, long right-angled USB-C cables don't cost a fortune. https://github.com/alvr-org/ALVR/wiki/ALVR-wired-setup-(ALVR-over-USB)

I'm at college right now, which is a 3 hour drive away from my home, where a server of mine is. I just have to ask my parents to turn it back on when the power goes out or it gets borked. I access it solely through RustDesk and Cloudflare Tunnels SSH (it's actually pretty cool, they have a web interface for it).

I have no car, so there's really no way to access it in case something catastrophic happens. I have to rely on hopes, prayers, and the power of a probably outdated Pop!_OS install. Totally doesn't stress me out I'll just say I like to live on the edge :^)

What's Reddit?

I don't know the specifics behind why the limit is 72 bytes, but that might be slightly tricky. My understanding of bcrypt is that it generates 2^salt different possible hashes for the same password, and when you want to test an input you have to hash the password 2^salt times to see if any match. So computation times would get very big if you're combining hashes

bcrypt has a maximum password length of 72 bytes.

https://en.wikipedia.org/wiki/Bcrypt#Maximum_password_length

Good to know! I suppose it makes sense for the smaller registries to be a little shadier.

While I do agree that this is bad, I'm a little confused—what does this have to do with dead internet theory? Doesn't that relate to users being bots?

I'm sure a lot of forks will pop up right around this time. I'll be less skeptical of them once I see actual commits made to the codebase instead of things like just changing the readme