I don't think it was intentional, the dev seemed to be struggling with health-related problems and possibly burnout. But yeah, definitely a depressing moment for an otherwise really cool project.

Honestly, thank you for demonstrating a clear limitation of how things currently work. Lemmy (and Kbin) probably should look into internal rate limiting on posts to avoid this.

I'm a bit naive on the subject, but perhaps there's a way to detect "over x amount of votes from over x amount of users from this instance"? and basically invalidate them?

So, to be clear, the story the article links to is specifically a case of local content that didn't actually federate. It was an accidental upload, he cancelled the post, it sat in storage, and even his admin was stumped about how to get it out.

I agree that with federation, it's a lot more messy. But, having provisions to delete things locally, and try to push out deletes across the network, is absolutely better than nothing.

The biggest issue I have is that there's really not much an admin can do at the moment if CSAM or some other horrific shit gets into pict-rs, short of using a tool to crawl through the database and use API calls to hackily delete things. Federation aside, at least make it easy for admins and mods to handle this on their home servers.

Did they, though? A bunch of other Fediverse platforms have supported this for literally years, to the point that Mastodon was the butt of jokes for breaking basic search functionality.

Having standard search that just works is a huge deal, and helps solve against the decentralized content discovery problem.

I'm pretty sure they mean respective to themselves and their own walled garden, but it definitely doesn't scan well.

Nobody is telling you to use it. This originally spun out of development of a messaging app just for Pixelfed, but evolved when the dev realized it could be made to work with any Fediverse account, not just his own server project.

An optimistic view is that it could end up opening the door for end-to-end encryption to come to private messages in Fediverse servers, over time.

Just cross your arms, smile wryly, and comment on how pathetic the Interviewer's pen is. Cheap material, runny ink, a grip that's painful to hold. Wish him good luck in taking notes on subsequent interviews.

Then lean in, and say "But, you know? I've got a premium writing utensil. It's crafted in the Netherlands by a Space Age engineering firm. It's designed to fit comfortably between your fingers. And the Indian ink that runs through it glistens and glides smoothly through a specially crafted tip."

Pull out a business card with absolutely beautiful handwriting on it. Just as he expresses surprise and interest, sigh and say "But... It's really not for you. It's really more of a thing for your boss, or your boss's boss."

Start getting up to leave, and wait for him to come running after you.

He is a bit bombastic, and has a habit of biting off more than he can chew sometimes. I think these side-projects are ultimately useful, though, and probably help fend off boredom or burnout. Maybe he gets better at coding and design through doing that, I dunno.

Regardless, he's continues to do a lot of great development work.

There's nothing wrong with having good third-party tools, that was not my point. db0 in particular has done some amazing, amazing work.

What's fucked, however, is having a project:

• whose core infrastructure only offers the most threadbare tools
• there's zero consideration from development on privacy, user safety, or basic controls to handle when shit hits the bed
• the devs are stone silent when waves of CSAM crash through instances
• they openly mock people or say they're "too busy to do this" when it comes to meeting the most basic expectations of how a social platform ought to work.

Like, this is not an attack on Lemmy itself, I think the platform can be a real force for good in the Fediverse. But let's be honest, this project is not going to live very long if nothing changes.

Basic things like having the ability to easily remove images from storage should be part of the core platform. The fact that this still isn't a thing even four years into the project is insane.

A few questions:

1. Why did you name it Lemmy?
2. What have some of the biggest challenges been in developing a Reddit-like community platform?
3. What's a big feature you hope to implement someday?

Technically, yes, you save metadata of all of those things. However: you are not a company that profits from vast amounts of data ingestion.

Hi, so I run Spectra, and would like to weigh in.

I wrote this piece a few years back about the content situation on PeerTube.

TL;DR, PeerTube has a significant problem with spam. I'm not just talking about spammy comments, although it has those, too. No, I'm talking about the videos. For example: there's an option within PeerTube that, when enabled, basically just automatically subscribes to every server that subscribes to yours. I let that setting run for a while, and connected to maybe a hundred random instances over time.

It was all garbage. Either you get far-right propaganda videos, actual nazi videos, or super random weird stuff of little value. Want a video in Hindi for a restaurant with a two-second video featuring a French TV commercial transcribed from VHS? How about that, mixed with thousands of random snippets of media that you will never care about or relate to?

A lot of PeerTube admins kind of informally got together and said: you know what, this is crap, no one is ever going to enjoy this. So, we connected our communities together. We have to do our research on which servers are good, and which ones just serve up bullshit. Good community stewardship, in this case, requires us to do our homework on which servers are worth following. Instead of following as many servers as possible, we're more inclined to check and see if the place is putting out original stuff, has decent guidelines, and isn't spouting hateful crap everywhere. To build community organically, we have to do so with intention.

The reason that you're not seeing your videos in any of the places you've listed is because their servers don't follow ours. This doesn't mean that your videos cannot be seen through federation - it's just that, in any of those places, no one is subscribed to you, and that server isn't subscribed to our server. So, your channel and videos aren't likely to show up there, unless somebody actively chooses to subscribe to you.

I agree that PeerTube is seriously lacking in some kind of Community Discovery feature, and would be greatly enhanced by it.

They probably thought they'd reach a greater number of people. They're probably right.

This is a situation that I think will get better in time. There's some really promising efforts involving Fediverse Enhancement Proposals, where multiple projects collaborate on shared ways of doing things. Some of these behaviors are getting studied and standardized by the larger SocialCG entity, as well.

There's also a lot of promising development behind a Fediverse Testing Suite. If we can develop a platform-agnostic testing system for people to build against, it will potentially become the new development standard, rather than optimizing for Mastodon and nothing else.

I take it you've never run a community instance. The problem is, laws vary by jurisdiction, and can have a very real effect on how you run your server when shit hits the fan.

We recently ran a story about a guy building his own Fediverse community and platform, who just happened to be a bit naive about the network. He's off in his corner, doing his own thing, people find his project and assume it's some kind of weird scraper. After disinformation came out about it, someone remote-loaded child pornography to his server, for the purpose of filling a report with the police.

The guy is based on Germany. Local jurisdiction requires one year of prison time minimum. It matters.

I've reviewed both your and @Nutomic's comments, your latest blog updates, and GitHub PR's, and added a section accordingly: https://wedistribute.org/2024/03/lemmy-image-problem/#giving-credit

Thank you for your hard work, and for taking necessary steps to improve something that is essential for instance operators.

You know, that's actually a really good point. Dansup tends to iterate on certain parts of his apps several times over, like how Pixelfed handles image uploads and filters prior to posting. It might be that just going with the simplest possible thing makes sense for right now, until a better approach can be devised.

Ostensibly, yes. However, as a company whose business model is primarily predicated on sale of personal data and analytics, this does create something of a conflict of interest, especially because of Meta's extensive involvement in surveillance capitalism.

Per the article, I really like Mike Macgirvin's stance of "I'll give you the bare minimum of data to make basic interactions work, but not one thing more."

Yeah, I feel like they'd probably get better mileage with something like Meilisearch, which is what Firefish uses for search.

Thing is, for federation to work, his team had to opt into it. The fact that his statuses and profile render natively in Mastodon and Akkoma are a pretty strong start.

I'd like to see Meta put their money where their mouths are, and finish the integration. I think we'll probably see that happen sooner rather than later.

The main thing is that their frontend was built for whatever APIs their own platform supported. You can theoretically adjust the code to use the Mastodon API instead, but it can actually be kind of a tedious process to change everything over.

It's probably easier to mod an existing frontend to look and act the way they want, or write a new one from scratch.

The other thing is, it's still a really new instance, and kind of started as a community experiment. While it looks like it could be a big next step for Pebble, they're probably more interested in testing the waters before doing any serious committed development on it.

Signal is great, but the core developer is inherently against federation with his own product. It's just one giant centralized service.

Sentry also did this by embracing the Business Source License. Technically, you can still get an MIT-licensed version, but it has to be more than two years old.

As a former employee that worked there during the days that Sentry really promoted itself being Open Source, it was disappointing to see. VC Funding and a growth obsession basically poisoned the well.

I think it's intended to be more like WhatsApp, in the sense that you use it for one-on-one chats or messaging small friend groups. I don't think it's a current goal to try to take on Matrix / Element.

Diaspora used to be great! The problem is that, as a project, it's been kind of rudderless in direction for a long time. It's been in maintenance mode for nearly a decade, where contributions are largely just random fixes and minor improvements sent in by volunteers.

Unfortunately, the old guard is very against adopting ActivityPub in any way, shape, or form. Historically, the project has always kind of put the expectation of federation compatibility on other platforms, rather than doing any work to collaborate with existing platforms or adopt existing standards. They can't even communicate with most of the Fediverse these days.

The project's future is kind of uncertain. They finally got a developer API put together, and work is happening on Account Migration. But, the platform is slowed down by years of cruft and technical debt.

Yeah. While I could take or leave Hubzilla's UI, this concept in particular was mind-blowing. The fact that it's like, 8 or 9 years old at this point, and the rest of the Fediverse barely supports it, is crazy to me.

Yeah, I agree. I think the important thing is "was the local content scrubbed?" Because at least if that was done, the place of origin no longer has it.

Federated deletes will always be imperfect, but I'd rather have them than not have them.

What might actually be interesting would be if someone could figure out this type of content negotiation: deletes get federated, some servers miss it. Maybe there's a way to get servers to check the cache and, if a corresponding origin value is no longer there, dump it?

Pixelfed is basically a federated alternative to Instagram that's been picking up a massive amount of momentum. The Groups feature is more or less federated Facebook Groups, with a bunch of robust tools built in, and will be compatible with Lemmy communities and Kbin magazines.

The reason this is significant is that it may prove to be the final push that makes groups a standard part of the Fediverse experience. Some of the biggest platforms in the space have lacked it, for years and years. This implementation could prove to be a really good blueprint of what the standard experience ought to provide, at a bare minimum.

That's actually a really cool incentive. Lemmy seems like a relatively easy target, too, since each node can help stabilize the overall network.

This entire comment made me laugh out loud, thank you. You have a beautiful way with words.

Look, I've been on this network pretty much since it began 15 years ago. I did community management for Diaspora back in the day, regularly kept in contact with many people who ended up starting their own platforms, and have had a pretty good finger on the pulse of the network's discourse during all that time.

The general attitude of people who have come here has largely been one of these:

1. Hey neat, an open source thing I can tinker with! And it federates!
2. Fuck the place I'm coming from, I never want to have to deal with it ever again.

There are a myriad of motivations behind how and why different parts of this space was built. For a lot of people, it's a place to build their own communities, hang out with a bunch of people into niche stuff, and generally just chill out and have fun. For others, it's a lifeline for their own marginalized groups, a way to find kindred spirits and support each other.

But, I guarantee that it sure as hell wasn't to build pipelines back to centralized corporate silos run by people trying to maximize profits.

Blocking at the instance level at least lets communities keep that crap out of our streams, and that's a feature of the network. Meta can implement AP all they want, and they'll probably connect to a decent amount of servers - but a lot of people can say "Fuck this." and choose never to connect to them in the first place.

They can go right on doing what they're doing, and their experience will be exactly like it was before Meta showed up. Being able to see and interact with who you want to, and filter out who you want to, is a fundamental feature of the network.

Hey, I'm the guy who started the .ml fediverse community. I started it with the Lemmy part of the network was young, and there weren't many instances yet. It's become a very active community, and I'm constantly amazed to see how much faster things move these days.

This has kind of been an ongoing conversation in some prior feature request discussions for Lemmy. One idea is that communities could consensually relay posts from one together, effectively creating a group containing Group Actors. This would probably cut down on duplicate content, but could create a larger surface vector for spam. But, I think it's an interesting idea.

I don't really have a full idea of what the best solution is. A Fediverse-specific instance similar to socialhub.activitypub.rocks could be a really interesting experiment, in that it would try to serve as a "Neutral Zone" between instances while sharing all kinds of news.

In the end, I don't really have much of a horse in this race. I think cutting down on duplication and redundant communities in favor of a more active shared space would probably have a lot of benefits, there's always going to be independent communities dedicated to the same theme on some far-off server. I'm not really interested in preventing anybody from starting their own.

I'd actually love to see something like this happen, as it seems to be something European governments and officials are embracing. To have something similar for the United States would be incredible.

I think at the moment, there's a real need for advocates, consultants, and vendors that can actually cater to government entities here. I would imagine there's probably some crazy data requirements needed for US Government Officials.

Yeah, I get that it's not that new, especially since it's a rebranded fork. But, as a fediverse project with its own brand and design sensibilities, it's relatively new, especially compared to Mastodon.

Unfortunately, I've used the Antennas for like two months, and they're janky. The "non-retroactive data fill" point that you make is only partially true, and seems to mostly apply to filters pertaining to a collection of users, or a collection of servers as a data source. It's a confusing UX papercut. Worse yet, my Fediverse Devs antenna example has been around for two months, and barely produces anything most of the time.

I don't believe that an "official branded app" is strictly necessary. What I meant was more in line with "Firefish could really use apps developed for its features specifically." Sadly, Misskey compatible apps continues to be a wasteland compared to the plethora of Mastodon ones.

The whole "the flagship instance is a sandbox test instance" is kind of a sometimes-true sometimes-not situation. It's definitely less stable now than it was two months ago, but that kind of messaging and expectations management didn't seem to really happen until sometime after the CalcKey migration.

And yeah, proper group / community federation with Lemmy is a huge deal! I'm looking forward to what Pixelfed is doing with Groups, as they look somewhat similar, and aim to maximize compatibility while providing good management tools.

It actually does have that nowadays, it's just that the feature requires Elasticsearch to work, which is one extra piece of infra for admins to worry about.

Wow, thanks for the great in-depth feedback! 🤩

Yeah, there's definitely other areas I could've delved into, like Public Clips or MFM or pagebuilding. These in-depth reviews are challenging, due to trying to strike a balance between features and actually getting something published. Most articles of this nature takes me weeks, sometimes even a month or more.

The UI definitely has a learning curve, too, but as a veteran Fedi user, it suited me just fine. I've dealt with far, far worse 😂

The instability really bums me out. I'd like to think that things are slowly improving, but the lack of transparency (and frankly useless error messages) make it really hard to triage where the problem is and forge a path forward. The lead dev has also been sick recently, and suddenly is not very active online.

Finally, I think Firefish takes part in a long tradition of Misskey forks, where a half dozen systems all branch off of each other. It's a shame that more of them don't collaborate on the same platform, leaving many devs to cherry-pick across forks. I wonder sometimes whether this hurts development more than it helps.

I agree that the problem isn't with the Fediverse itself, any more than it is with email, usenet, encrypted messengers, etc.

The thing is, it's a problem that affects the network. While "block and move on" is a reasonable strategy for getting that crap out of your own instance's feeds, the real meat and potatoes of the issue have to do with legal and legislative repercussions. If an admin comes across this stuff, they have a legal obligation to report it, in most jurisdictions. In fact, the EARN IT and STOP CSAM acts that politicians are trying to push through Congress are likely to make companies overreact to any potential penalty that could come from accidental cross-pollination of CSAM between servers.

Unfortunately, this thing becomes a whole lot messier when an instance discovers cached CSAM after the fact. There was a Mastodon instance that was recently taken down without any turnaround time given to the admin to look into it, the hosting company was just ordered to comply with a CSAM request that basically said "This server has child porn on it."

Also, regardless of whether you report it or block it and pretend you never saw anything, that doesn't change the fact that it's still happening. At the very least, having tooling to make the reporting easier would probably be a big boon to knocking those servers off the network.

I think there's a balance to be struck between "good defaults" and "customize to your heart's content."

Emissary is very much in line with some of my own pipe dreams regarding Fediverse / IndieWeb platforms, but it's still very young as a project. I think the best thing they could probably do is ship bundles of templates as different experiences, that are easy to install right out the gate.

Want a bog-standard microblogging system? Go for it. Want something more like Lemmy? No problem. Want to just build something yourself from scratch? Here's the docs.

I think what excites me about this is that it could be a tremendous development tool for people looking to mock up new ideas for apps and platforms, while sitting in top of ActivityPub and offering actual functionality. The Music project the lead dev is working on already looks great in less than two weeks of development, and aims to be compatible with Funkwhale.

I dunno, federated communication systems have a very different utility than lemonade stands. 😅

While I think shareholders can be a driving factor, I see it way more often with VC-funded companies. The "2.5x year over year" growth mantra that places like YCombinator stipulate have disastrous effects on small tech companies. Often, these startups have an incentive to keep taking additional funding rounds, which appears to tighten the grip the VC has over them.

Try growing the next Microsoft or Google or Amazon out of that model. I'm not convinced that it's possible. At least if you bootstrap your own company, you don't have the same binding obligations...even if it takes way longer to get to a place that's self-sustaining.

No, you're good, and we're mostly on the same page! My general expectation is that your server tries its best, maybe there are still copies out there, but you shrug and say "eh, I wiped my stuff locally, good enough".

But yeah, I agree that once something leaves your server in terms of the passage of data, there are no guarantees. And I do agree that significant structural changes are necessary and important for the network's continued evolution!