I joined during the API blackouts, wanted to show some solidarity by not using Reddit during that time.

Then stuck around because I liked it. Though I'm still on Reddit too.

I'm not on Lemmy though, but I know you mean the threadiverse in general. I'm on Mbin.

I honestly haven't checked if it all works — the AI kept telling me it compiled and the endpoints returned 200, so I'm gonna take its word for it.

AI literacy 0/10

No, the AI telling you it works is NOT reliable. Test it out yourself, or if you can't be bothered, don't post it. Don't make others do the testing for you. Unless the AI showed you proof, it's not a reliable source here.

Right, I forgot the Piefed system is about votes you give out while the Reddit system is the opposite, votes you receive. I mean I did remember, but I forgot to use my critical reasoning there and notice the mismatch.

Mbin does have karma however. So it's not like it doesn't exist at all on the fediverse. Your karma is 1645 for me btw, though obviously depends on the posts of yours that have been federated to my instance.

Also, this isn't entirely correct about Piefed. It does actually have a karma system, it's just based on your up-/downvote ratio, not a number that ticks up endlessly. Some instances like piefed.social also have a hidden karma system that penalizes you for what it considers bad behavior iirc.

edit: the piefed part isn't actually relevant

Mbin does have karma however. So it's not like it doesn't exist at all on the fediverse.

Also, this isn't entirely correct about Piefed. It does actually have a karma system, it's just based on your up-/downvote ratio, not a number that ticks up endlessly. Some instances like piefed.social also have a hidden karma system that penalizes you for what it considers bad behavior iirc.

(edit: that latter paragraph isn't really relevant)

If any Lemmy devs look here, Mbin adds the image to the attachments array and uses a custom property (source) to add the url directly to the post instead of url (in addition to the entry in the attachments array which Mbin adds too). I don't know how you're getting the url, but I assume one of those two is the issue. Both can be fixed by the Mbin devs imo, but if the attachment array contents are the issue, I do recommend making Lemmy more robust there as the problem won't end with Mbin in the future. Other projects might also add additional objects to that array.

I haven't taken a look at the Lemmy source code, but judging by the JSON-LD, I think this could easily be fixed by ordering the items in the attachment array so the link comes first. Lemmy seems to just take the first item there and then abort if it's an image not a link.

Depending on how Lemmy is implemented, it might also look for the url property. I don't understand why Mbin is using source instead of url. If Lemmy falls back to the url property, the devs could add that to posts.

Not a dev btw, just trying to provide more context.

Like I said - BASED ON WHAT DATA?

This data for example: https://gs.statcounter.com/...

Latest stats:
Windows - 62.06%
Unknown - 18.74%
OS X - 10.49%
macOS - 4.21%
Linux - 3.09%
Chrome OS - 1.4%

There's also the Steam Hardware Survey, but that's specifically gaming devices, so not the most reliable overall: https://store.steampowered.com/...

Windows: 93.85% Linux: 3.99% OSX: 2.16%

If you've manually verified the tests, then you prove that the tests are a proof

Yes, I was objecting to

Nothing an LLM will output is ever a proof.

If you prove the tests are a proof, the LLM will have still output proof. It's still what wrote the tests.

It didn't prove anything on its own of course, but that's not what I was objecting to.

The Right to Erasure would like a word with you there.

If the AI writes unit tests that function correctly and do what they're meant to, I don't see how that isn't proof for the code the unit test tests.

Lemmy doesn't have that? That does sound like something they should change then.

Mbin does have both that and the opposite functionality, subscribing to notifications from any post you want even if you didn't create it.

For people who only go off headlines and comments and don't read the article, here's the important bits:

• This only affects some users, not everyone
• The issue is that comment sections are disabled and video descriptions claim to be empty
• The issue goes away with a refresh, you don't need to turn off the adblocker, though you'll have to refresh every video
• It affects all videos
• The link to adblockers is due to everyone who is reporting the issue being a user of an adblocker, and turning it off fixing the issue permanently until it's turned on again
• There's no specific browser or adblocker mentioned in the article
• It affects Premium users too

edit: added "the article" after "don't read" to clarify

For anyone reading this but not the article, the issue is that the AI interprets the query literally as a command to disregard the query. So Google's AI does exactly that. The actual search works as normal, and Google didn't ban or otherwise censor the word.

I'm terrible at drawing, so I in fact cannot.

But I'll take this as an opportunity to experiment.

Reddit didn't remove it, the moderators did. On Shreddit, there's a note:

Sorry, this post has been removed by the moderators of r/YouShouldKnow.

It's still available on your profile, I can still see it there. So clearly Reddit didn't do shit.

Edit: While I'm looking at shreddit, looking at the rules in the sidebar, I think it was removed due to rule 3:

YSKs regarding Reddit, Facebook, Twitter or any other social media are NOT ALLOWED.

You should always look at a community's rules first before posting, imo this removal is justified.

It's called projection. You accuse an enemy of something that actually applies to you or your team. Heavily used by people on the far right. People also refer to it as "every accusation is a confession".

Like when Musk called that diver a pedophile, when Musk is the actual pedophile.

That's why the accusation seems odd since the antichrist is more likely to come from the Republican party than some random activist. Because this is actually about the Republican party, not Greta.

Worth noting that while the rest of the world uses rectangles, it's not always the same dimensions.

The Swiss flag, for example, is explicitly square. If it's not square, it's actually the maritime version specifically.

In other western countries, the state asks for taxes so it can use them to provide services to its citizen. In my country, states actually lower taxes when they've had too much net income in a year. Because why ask for so much money if you don't actually need it.

In the US, it sounds like taxes are seen as a tribute to the ruling elite, with anything given back to the people seen as evil and undeserved.

I'm a bit confused by comments on this topic. Do sovereign countries not have the right anymore to decide their own laws and issue punishment when they're not followed?

Like, they obviously can't enforce these fines. This article says as much. The fines can't be enforced, but if 4chan ignores them, that opens the door for other measures like delisting the site from search engines or blocking access to it from the UK (these two examples are taken from the article). Which are fair measures imo.

Like, to the people saying UK can't do laws which apply to services which are merely accessible in the UK and have no physical presence there, do you also apply this logic to the GDPR, which works the same way? The US has these laws too, like COPPA iirc. It's not really something the UK came up with, it's a bit of a standard with laws like this as far as I know.