Also, ML is just statistics and calculus.

There was an unfortunate email glitch where if your username was already present in the instance, it would silently deny your application, and the beehaw admins would be none the wiser. See https://beehaw.org/post/562922

Additionally, there has been some issues with the beehaw's email provider that has recently been fixed and the backlog has started to decrease. See https://beehaw.org/post/604680

Can you explain what you mean by bad experience? I can see how the experience on beehaw would be different from other instances, but what is explicitly bad about it?

1. The account registration process exists to weed out bots. We're not the only ones to implement this kind of sign up. An essay really isn't required, unless you misunderstood the purpose of the sign up.
2. We haven't left the fediverse, posts and comments to and from beehaw still flow to the vast majority of instances, and we do wish to rejoin these two specific instances at a later date once we have the right processes and tools to work though the problems we encountered.
3. If the admins were cocky or snobby, they would have defederated without any form of announcement or transparency on what was being done and why.

Unfortunately, the inconvenience is something of a catch 22. Do we allow everything through for the sake of convenience? What happens when extreme content that is NSFL gets posted? What happens when illegal content is federated, or hate speech that indicates action will be taken is made? What happens when you observe a pattern of this behavior from a common source? Content must be moderated for things to be "safe" and the rate that unsafe, nonaligned content was coming in wasn't sustainable.

Choosing to defederate wasn't taken lightly and it was done reluctantly. It was discussed for two days after observing systemic effects from those instances and after reaching out to the instance admins for alternatives.

I see you're posting not from a beehaw account, which means you likely haven't seen @Gaywallet@beehaw.org 's post on what it is to be a community and the framework to get there. This posts may help you understand this instances stance on things and what our instances users are hoping for is to build.

All in all, sorry you're not happy, but we're being careful for our community.

Unfortunately, defederating means the cord has been cut. This means we still have what was previously been posted, but all future content is bidirectionally blocked.

This is true, except for one element:

Fediverse should mean a user of any instance should be able to use any community the instance elects to federate with. Lemmy is open by design, but instances can just as easily switch that feature off and go to a allowlist method.

A commonly missed element with federation is that you federate with who you trust since you essentially mirror their content. It's less apparent with the lemmy migration, but mastodon used to caution its users to "join an instance that aligns with your preferences" for this reason.

Federation is really a philosophy about mutual trust, just like how email providers can block messages by user, instance, or domain.

Trust me, there's likely more gating present than you're aware of. Maybe not at lemmy.world (which as of this post is only blocking one site for reasons I won't mention), but this can get dark pretty quick if you leave things completely open.

If this is true, then it's not good news

https://kbin.social/m/RedditMigration/t/34112/Heads-up-Reddit-is-quietly-restoring-deleted-AND-overwritten-posts-and

List of hackernews folk are saying first changes were reversed https://news.ycombinator.com/item?id=36354850

Hey everyone we're trying to keep the reddit threads centralized in technology in beehaw. I'm not locking this one because there's a lot of discussion, but consider moving the chat over to https://beehaw.org/post/576904

Content prior to the defederation will stick around, but any updates and comments will no longer be synced. So what you're seeing now in our instance is a snapshot of those posts for what was present prior to the defederation.

It's a good point, and could alleviate things a bit, but its not easy and still would have problems. The root of it is that there's a rate limit for normal accounts in the new data api, so you would quickly get a message saying you've done too much in one minute and basically stop working. This also assumes the app can move over to a user provided authentication within the time window of one month. Given many of the big apps use support servers thar help manage the api calls (such as apollo's), this can be a big ask depending on how it was coded, especially when they didn't need to in the past and have built up around the lack of a need for oath for a decade. There's also an education process to get users to request and wire in the oath (could be done well, but it would be much rougher than today).

Additionally, there is still the censorship of nsfw posts, which enforces a walled garden, so third party apps would always be inferior. This sets a precedent more so than a problem, because next up could be whole subreddits being omitted from the api because of future reason here.

Also there's legal concerns it might bring up like stated by andrew