No-one cares what Russia or China thinks here. Germany? I mean, sure, but this is also a complication for any regulatory bodies trying to police social media sites. As “Lemmy” or “Piefed”, as you know, are not singular entities.

People living in Russia or China might care.

UK might be much more difficult, btw. They now ban all porn without identity checks. So if you host a lemmy/piefed instance that’s accessible in the UK you will need to delete all adult content that makes it to your instance, if you don’t want to violate UK law.

People would just clone the communities on other instances and rebuild.

Cloning communities isn’t quite that easy. Were you present when feddit.de went down? Their communities didn’t vanish. The replications are still up on all other instances, and you can still post there. There’s no indication to a casual user that the instance hosting the communities is down and thus federation doesn’t work. To the users it just looks like participation dropped like a rock with no obvious reason.

The communities were cloned onto a new instance (IIRC feddit.org) but even up to now, people keep posting to the old now-unfederated communities.

Btw: that’s another quite critical issue: Lemmy lacks any and all migration tools. Can’t migrate an instance to a new URL, can’t migrate users or communities to other instances. All you can do is scrap all you had and start fresh.

Sure, you’re right there - but an instance that kept having problems with removing CSAM would find itself defederated.

Depends… Imagine it also contains some of the most relevant communities and defederating would mean you lose users. That’s not such an easy decision any more. Also, at that point hosting would likely be so expensive that for-profit instances would emerge, and for those defederating an important community wouldn’t be such an easy choice either.

But it’s not only CSAM. For example, there’s illegal speech in quite a few parts of the world. In Germany, for example, a lot of nazi-related stuff is illegal. In russia or china some regime-critical speech is illegal. I wouldn’t be too surprised if the US also joins this club sometime in the near future.

Actually, if you are a non US citizen and you and you want to travel to the USA, it’s already troublesome if you are hosting a website with anti-Trump content.

That kind of stuff is unlikely to be deleted on the original instance if that instance isn’t hosted in the same country.

Yes, so not financial. You seemed to be implying it was financial.

Sorry if that came across. I said lemm.ee was shutdown because of the scaling issue. I could have been more clear with that I meant the moderation scaling issues.

Wasn’t trying to convince anyone of anything. Just offering a reality check.

Lemmy vs Reddit is like this meme where the one side says “I hate you” and the other side says “Who are you?” (or was it “I don’t even think of you”, can’t remember).

Lemmy is cool. It being small has benefits and I like the political direction here much more than on Reddit. I like that most people I interact with on Lemmy genuinely are humans. On Reddit, that’s much more difficult to be certain of.

But Lemmy is not Reddit, it’s not a Reddit alternative, it’s not even a Reddit competitor. It’s a nice little niche forum, a little anti-capitalist experiment, that kinda copied the UI and UX of old Reddit. That’s totally fine and it’s got it’s value. Otherwise I wouldn’t have >3000 comments on this platform.

But it’s a factor of 100 000x off of being on the radar of Spez and his crew.

Eh, if the original instance removes the CSAM - the ban and removal federates out to everywhere else, so this isn’t always true.

But if it doesn’t, then other instances removing the content on their side doesn’t federate. So you can either trust every instance that you federate with with your legal security, or you will have to moderate everything yourself as well, just in case someone missed something.

Down the line, the answer here would be for the federative structure to change so that an instance only hosts its own local content, and doesn’t need duplicate content viewed from external instances.

This would be extremely important, but I don’t know if such a low level conceptual change can still be performed with a reasonable amount of work. Remember, for such a change you need to get every instance on board. That would be difficult now, and only more difficult later.

Tbh, it would have been much smarter if the setup would be basically a bunch of independent phpBB-like boards with federated single-sign-on and an app that transparently connects you to whatever instance hosts the content you are looking at.

That’s not why lemm.ee closed down. It wasn’t financial.

No, it was specifically because of the moderation issue: https://lemmy.ca/post/45390962

Because Lemmy, to this day, doesn’t do what Reddit does. Yes, the UI is similar, but there’s two big downsides to Reddit. One that’s important now, and one that’s important later.

• Lemmy is tiny. Like, really small. The Linus Tech Tips form and the Crackberry forum each have more users and more activity than all of Lemmy combined. That means, you can talk about general things of Lemmy, like e.g. US politics, but there aren’t a lot of niche communities. On Reddit I can post a photo with some weird electronics component from the 60s and within minutes someone will post an answer identifying the component and telling me where to buy a replacement. On Lemmy, a corresponding community doesn’t even exist.
• Lemmy scales terribly. Every instance holds a copy of all data that was ever posted in any community that any user on that instance ever subscribed to. That has two very negative effects:
  • Storage requirements are insane. Since most traffic is in big communities and most users will subscribe to the big communities, most instances need to store a copy of almost all of Lemmy. If Lemmy were to ever get to the size of Reddit, every instance would have to store data in the order of magnitude of all of Reddit. Imagine small hobby admins having to host data in the region of Petabytes or Exabytes. Nobody can afford that.
  • Admin work is insane. Since every instance holds a full, independent copy and doesn’t only cache, they are legally responsible for the content and have to moderate it. So if someone posts e.g. illegal pornography on one instance and it’s federated to another instance, the admin of the second instance needs to delete it or face legal consequences. That means, instead of the mods or admins of the original community/instance being solely responsible for keeping their stuff clean, everyone is responsible for everything and the same work needs to be done hundreds of times, once per instance.

This horrible scalability means that right now instances are getting close to their limits (see e.g. lemm.ee closing down exactly due to these reasons).

Lemmy has 40-50k monthly active users. Reddit has 5.16 billion monthly active users, so about 100 000x. If everyone on Reddit were to move over to Lemmy, Lemmy would be done. Just one day of Reddit-level traffic would be enough to jam up the history of Lemmy content so much that nobody could ever afford hosting a Lemmy instance again.

Or sometimes its a job they legally need to have but where they really don’t want quality work (e.g. support).

Found the esoteric programmer!

How much would it cost to just use “scavenged” memory?

Also, comparing indian prices to US prices ist’t exactly fair either.

I’d be very surprised if that was able to damage your speakers. If it could, then so could any sound you play. It would have to be an exceptionally bad design if playing a youtube video could destroy your speakers.

And with a command like ffmpeg a wrong bot command might just as well overwrite your source file.

So like

# Stupid shitty piece of fucking useless waste of compute resources! Do your fucking damn job, you useless clanker!

divideMp4IntoNSegmentsOfLengthT() {
}

Explicit enough?

I am so surprised how many of the predictions kinda came true.

It didn’t happen exactly how he tought, obviously, but a lot is actually happening.

Lemmy fetches everything that has ever been posted in any community that any user on that instance is subscribed to and keeps it indefinitely.

Since most activity happens in big communities that most people are subscribed to, most instances keep full, persistent copies of most things that were ever posted to lemmy.

That’s why Lemmy scales so badly. If Lemmy was the size of Reddit, every instance would have to have storage capacity in the same order of magnitude as all of Reddit itself.

The problem only gets worse with time, since all that has been posted still remains.

The total replication also means that the copies need to be moderated by every instance individually, since every instance stores a copy of everything. So if e.g. someone posts illegal content on another instance and your instance stores a replica, you are just as legally liable for that illegal content as the original instance. Thus you have to moderate everything that runs over your instance.

Moderation effort is thus also replicated across all instances.

That bad scaling in storage and moderation is btw the reason why e.g. lemm.ee shut down. It was just too much cost and work to keep the instance running.

Same as MicroSD

I do agree with this. I’d love to know more about the software, especially about how long it’s going to be supported. I hope they release more info closer to release.

deleted by creator

That’s exactly what the article is about. Please read it, it’s good.

Tbh, I would classify everything you do regularly at work as your daily work.

The point remains the same: You need to know all these languages. You cannot say “I’m just doing DB stuff today, so I can totally forget QML and relearn it tomorrow when I have a GUI ticket.”

Tbh, I’ve worked with self-professed “superstar programmers” so far. In every case their cockiness, ther “I’m always right” attitude and their inability to receive criticism or to adjust in any way to the team lead to major issues.

These “superstar programmers” are much more likely to produce live bugs (because they don’t listen to anyone, don’t follow procedure like code reviews and stuff and after a time nobody wants to invest time into reviewing their code properly because it always ends in a fight). They are also really harmful to team cohesion. Their behavior usually ends with most of the team quitting unless they are fired before that happens.

Programming is teamwork. It by definition needs to be unless you work alone. Someone who can’t work well in a team is a bad programmer.

We have a superstar programmer in our sister team. Both teams work on the same project but in different sections, but there’s a core area where we overlap. That guy recently caused a massive 5h outage of our customer-facing website, which is our main access point for much of the business. We have customers in the order of 100 million that couldn’t access the website.

What had appened is that our superstar programmer worked in the middle of the night when nobody was online. He made a non-urgent but critical change and pushed to production without code review or testing by the QAs, because, you know, he’s a superstar and he doesn’t need to follow processes. Our process for a change like that would have been that he needs a code review from his team and from our team (since the change affects us as well), and then it would have to be tested by their and our QAs in the preview env before and on production after the push.

There’s around 20 domains for our website for different countries and different client companies. Our superstar only tested one of them after the push and that one was working, so he didn’t check the rest. Turns out he made a simple typo and all other domains weren’t accessible, until the rest of the team members came online at ~9 in the morning and noticed the massive outage. We then spent two hours trying to troubleshoot what exactly went wrong until someone found and debugged his merge request, because of course he didn’t even inform anyone that he did the change. (He didn’t even create a ticket for it, even though our processes require that.)

Superstar programmers are great at contests but terrible workers.

It’s pretty much the same as why you wouldn’t hire an artisan instagram bricklayer influencer to work on a block of flats.

So you built your own location tracking service for privacy, and then you post your tracking online? ;)

Seriously though, well done, quite impressive!