The point of YouTube is that all of the content is posted by users, not by an editorial staff. That's what makes it "social media". If YouTube isn't social media, neither are Tik-tok, Instagram, etc.

It's a bit more than that. The YouTube video feed is also generated based on your preferences and what you watch - just like the feeds on other social media sites designed to keep you scrolling and scrolling and on the site.

The whole "liking and subscribing" doesn't do much other than help YouTube refine the content it shows you to personalize your feed even more.

That makes it much more of social media than just a content delivery platform.

If instead its front page just showed you the popular videos and if you wanted to watch specific videos you went to a special "latest video" feed that just pulled the latest uploads from your saved content creators, then it's less likely to be a social media site since it's just showing content that you've asked it to show you and no algorithm is behind the scenes trying to keep you hooked.

It's really the whole "you have a feed tailed to your preferences" part that really defines social media. Facebook has its news feed that pulls from your friends and related people. YouTUbe has the main page, TikTok has you endlessly scrolling the same kind of thing.

The only social media site without such a thing would be Mastodon - the only algorithmic feeds are what's popular on the server, what's popular on what users on the servers viewed, and what's the latest posts on the people you subscribed to. Lesser known people would be based on your server (you picked a server that represents your primary interest, so other people on that server would fetch stuff you might not otherwise see).

And most of it is because the Republicans or whatever your local right wing political party equivalent is are making a huge stink about porn because the economy in every country they are running its collapsing so they need more moral panics to make their voters feel like they're getting their money's worth. Because they're losing all their money voting in these Jokers.

No, they are making a big stink about porn because if you've read Project 2025, the goal is to eliminate pornography.

Everything Trump has done so far (despite claiming to have never read Project 2025, not knowing anyone who wrote it (who is/was his senior advisors), etc), has been documented in that document. It was only a matter of time before they would get around to banning pornography (which is like the #2 action after eliminating LGBTQ).

And even though the Supreme Court has ruled that pornography in the past is a speech thing, expect a nice end-run around the court, or even a new ruling overturning such things (which would be ideal to weaking first amendment rulings and protections).

This is a great example of why one shouldn't consider anything put on the Internet anywhere to truly be secure.

Why would anyone consider it ever secure?

There's been at least a 50 year old adage that basically says "never put online anything you don't want to see as the headline on the New York Times". Or basically never post anything private online. And this was during the rise of early online services that the vast majority of people didn't have access to, given computers and modems were a rarity.

It's only grown in truth since then, even as people try to forget it. (Zuckerberg realized that "privacy controls" were the only way he could get uses to open up about their private lives and thus become more valuable assets).

Just assume everything is public. It may be hidden by some paywall or other thing, but it's still public.

It's actually less Windows 11's fault and more the game's fault. It's just that previous versions of Windows let the game get away with bad practices like use-after-free, or assuming an allocated page will be zeroed (it would be if it's a new allocation - for security reasons, but not if a page is reused by the same process - it's just things like Windows might tend to give you a reused page rather than incur the overhead of having to visit the memory manager for a new page). Some uninitialized data structures got corrupted in this way because they were relying on the fact that somehow one part was zeroed.

Several bugs were found of that class. Proton could easily result in the same issues since things like that are also not guaranteed by the OS.

So it is less compatible simply because it doesn't retain the same bugs. Granted, this wasn't like Windows 95 where Microsoft needed to make sure applications ran just as they did on Windows 3.1 or DOS and they had to write compatibility shims to get around them. Though I think Microsoft also said they dumped all those shims as cruft that was really adding bugs to the code.

I don't see it as a vendor leak. It was chaining two flaws that were pretty obvious.

The first flaw was an authentication bypass - which was so primitive it's a wonder why it's a 2025 CVE (it relied on "referer" headers).

The second flaw is related to deserialization of data, and anyone who does this knows it's something extremely tricky and is very insecure. Deserialization of any object is an inherently risky task, and it's something that's stymied experienced developers for 50+ years (back when it was a simpler "parse this file", that later grew into "handle this network protocol").

The real issue ended up being how the serialized data managed to cross the trust boundary from untrusted input to trusted input, get deserialized and then treated as trusted data. And given serialization of objects, well, you ended up with code being serialized in there (.NET intermediate runtime, actually). Kinda-sorta like Shell Shock all over again.

They were two independent flaws, the serialization one being a bigger more difficult one to handle because it was a fundamental thing in SharePoint, but likely safe because you needed to be an authenticated user to do the operation.

Thus, it doesn't seem like a stretch that someone who discovers an authentication bypass suddenly the second flaw becomes much more serious.

Of course, Microsoft is also going to use this to force more M365 sales because which SharePoint implementation wasn't vulnerable to this? Yes, cloud SharePoint was long fixed and patched. It was only on-prem installs of SharePoint that were affected.

If the CEO is a narcissist, he should exploit the situation, get paid to go on talk shows as an "anti-cheating" proponent that learned his lesson, maybe even a privacy advocate, then start a private investigation company or something. That doesn't seem to be his angle.

Except his wife will get an even bigger payout. He had a wife and two kids, and this very public blowout meant basically a huge divorce settlement with lots of money in child and spousal support.

The reason people are happy about it is the same reason people celebrated when Ashley Madison got hacked and their entire customer database leaked. The public just likes to see cheaters exposed, especially from the high and mighty.

PHB "logic". Instead of DBAN just the HDDs and SSDs, they physically ground up perfectly good servers and the drives rather than wiping and selling them.

I show you two drives. One has been DBAN and one has not. If the one that has not been erased leaks out, it means millions in fines. You have time only to erase one drive. Which do you pick?

Shredding a drive guarantees the data is destroyed. Running DBAN does not - because when you have a pallet of drives to get rid of, it's entirely possible for one or two of them or more to not get erased. And short of checking all of them, you cannot tell if there are unerased drives in the pile.

Shredding the drive, is easy to tell if you have missed a drive.

Erasing the disk works, but it's entirely possible to miss a disk or skip one if one happens to be distracted at the right moment. And checking/verifying/redoing the erasure is often not worth the time or effort. Whatever remaining value the disk has is likely far lower than the time needed to verify it's been wiped, that all drives in the batch have been wiped, or to redo the wiping process just in case. Just toss it into the shredder and be absolutely certain the data is destroyed.

The only people who can do it would be volunteer places like FreeGeek, and those places typically don't offer a data destruction guarantee because they also realize they can't have volunteers sitting around wiping disks and checking they've been wiped before moving them onto the floor.

The problem is real developers have to maintain such crap tools.

There have been a lot of fads over the years for "no experience coding" tools - from things like Visual Basic, JavaBeans and other RAD (rapid application development) tools.

The problem with all of them is they always somehow make it into business processes and end up having to be maintained well beyond their retirement because it's gone from something the boss wrote (or the boss's son/nephew/etc) over a weekend to a critical business process tool that breaks if you click on the wrong thing at the wrong time and needs hours of developer time to fix the damage it caused. And is so fragile no developer dares to fix the problem because who knows what other thing breaks it.

You know somewhere out there is some company still relying on VB6 applications with the source code long lost that's powering their critical back end services.

Vibe coded business applications are likely going to be next. The quicker they self-immolate and delete business data before they enter the business process, the better.

Trying to find out how much power they have. There is no way in Hell that they have 80 fully powered superchargers. They may have 80 superchargers installed, but I would be surprised to hear that they have even 45kW per charger.

Well, given people are likely going to be at the diner for a full hour eating, there is less need to give everyone 320kW that gets them charged up in 20 minutes. That's just silly. If you're going to spend an hour, then you tailor the fast charging to give you 80% in around an hour, though likely you need much less since people likely aren't coming in with depleted batteries. The largest battery in their cars is around 100kWh, so honestly, you probably could get away with 50kW chargers because most people will be around 20-30% and this should get them to 80% in an hour. You may have a few Cybertrucks there who need faster chargers so you can have excess capacity to handle the few.

But for the vast majority of patrons, 45-50kW is probably more than adequate.

Yeah, developers could see this coming a mile away. Heck, most people who saw the rise in ChatGPT and hallucinations could see that poisoning the AI was an attack.

People blindly following ChatGPT instructions on how to do something - you could see a potential attack where it gives you a command that would format your hard drive right then and there.

And "vibe coding" as the next big thing - seasoned developers could see it as a way to produce the next SQL injection or buffer overflow attack - attacks long thought to be well known and understood and no longer exist, are making a comeback. Vibe coders who don't understand what their tool is giving them, like ChatGPT users who blindly follow instructions, are going to get into trouble without knowing it.

We've just created more dancing pigs (or bunnies). It's how one of the first iPhone worms worked - people wanted to jailbreak their iPhones so badly (usually for pirated apps) that they installed all sorts of backdoors (norably, ssh) with default credentials (root/alpine).

Of course, because Arsenic (As) is directly below Phosphate (P) on the table.

The table is very useful for predicting chemical properties because a lot of things rollow - as you go down the table, the number of electrons in the valence shell stay the same, so many of the same chemical bonds can take place.

It's why Sodium-Ion batteries are a thing because they behave just like the lithium ions in a Lithium-Ion battery. Or why the Radium Girls lost all their teeth and had significantly weaker bones - because the body mistakes Radium for Calcium.

Or, more contemporary, PFAS. In PFAS, instead of carbon-hydrogen bonds, you have carbon-florine bonds (Remember hydrogen is actually supposed to be before Helium on the table, so its chemical properties can follow that group too). The body mistakes long chain carbon-florine molecules as their hydrocarbon counterparts (fatty acids, lipids, etc). Of course, these molecules are "broken" because they are extremely inert and thus don't really react in the body, which screws everything up - the body can't tell the difference between hydrocarbons and florocarbons as they do look very similar, just hydrocarbons react.

It's why the periodic table is one of the fundamentals of chemistry - many reactions you do have counterparts with other elements. Sure, things do change - the atomic size is different, and the bonding energy is very different, but the properties are similar enough that if you can do it with one element, it's possible with another. It's just that the result might not be stable if the bonding energy is so low it can't stay bonded, or that the size of the atom might make a huge difference if the molecule needs to fold. But the theory is there on how it should behave.

Just make it so if you do this, you lose, and the other side gets attorney's fees.

The judge can make it so the client doesn't pay the fees, but the attorney's do. The client still loses their case, but now they have a case against their lawyers for malpractice.

Any firm who gets known for losing cases because they use AI will likely go out of business.

And firms that use AI will want to double check their submissions because they don't want to lose their case over something trivial. And have to deal with malpractice lawsuits.

I was using the Archive to listen to BBC Radio shows (thanks for paying your license fees UK slashdotters!), but they got taken down. If it's an official library now, can they put those back up? It won't affect me so much since I snagged all the Finnemore and Mitchell I needed through even less legal means (thanks again for paying those license fees), but more people should be able to enjoy The Unbelievable Truth without having to deal with YouTube's added ads and general BS.

The Internet Archive was always a library - they got that designation well over a decade ago, and it gives them a few rights with respect to the DMCA.

The federal depository designation means its a library that can receive government records and make them available to the public. So you can go to one and get access to Supreme Court decisions, for example. And if it isn't digital, they can get a loan of the document and make it available. The Internet Archive is one of the few with a digitization program, so it also means they can request print-only government documents in any library in the system and digitize it and make it available. Sure, the government has its own program but it's often backlogged and other stuff.

The exciting bit might be to get documents from NASA that have be manually requested and scanned - which the Internet Archive might be able to do faster and cheaper and make it more widely available.

You shouldn't be running random scripts at all.

I hate this new trend for free/open source software where the install instructions are to curl some script and pipe it to bash. Or more fun, sudo bash. Because who knows what the script really contains, and a supply chain attack on that script easily compromises everyone who tries to install that software.

The question is if the studios demand it or if the test audiences demanded it.

Movies have test screenings all the time, and the original director's vision and script might get completely tossed because the test audiences hated it, requiring a last minute rewrite and re-shoot that changes the ending.

Studios demanding happy endings means studios have a chance to fix it at the writing stage before the movie is shot (this is MUCH cheaper). Test audiences only see the mostly-finished movie, so if it movie test screens bad, then you have to do a last minute change. Usually directors get a chance to fix the movie but sometimes they also have a right to refuse to do so. Of course, if a movie does bad because the director refused to make changes after a test screening, that director's chances of another movie go down significantly.