Kisai

Why not? Eminent domain is a thing.

Exactly.
 
There's only three appropriate solutions:
1. If a game is withdrawn from sale. It must become de-facto public domain.
2. If a game remains for sale but it's only the online component is withdrawn, they must open source the "server" component of the game, and patch the game so that the servers it communicates can be defined by the user.
3. If a game is a MMO, and a subscription, or microtransactions are used to secure "property" within the MMO, then a shutdown of the MMO must result in a refund of ALL property transactions back to the customer.
 
For example, Final Fantasy XIV, you can only secure your property (house) in the game if you remain subscribed. If Square Enix one day decides to shut the whole thing down, and you've been subscribed to it for 10 years, you should be given a refund of $1800 for the destruction of your property.
 
Games like Fortnite, if it's ever shutdown, should have every single Vbucks refunded.
 
Gacha-driven games like Genshin Impact, likewise.
 
The company only gets to keep subscription or microtransaction money if there is no attachment to property in the game. So if you bought a mount? You should get a refund. If you bought a skin? You should get a refund. If you bought anything that persists as an item, you should get a refund.
 
Square Enix, has shut down numerous microtransaction heavy games, and they're not the only ones. They run a game for 6 months or 2 years and then shut it down. For what? A tax write off? 
 
 
 

This unfortunately is going to open the floodgates. Like, sure, if you're cool with doing this, fine. But you start doing this too often what is going to happen is people are just going to follow Linus around when they see him in public and harass him.
 

There will be a steamdeck 2, and then there will be no steamdeck 3 because Valve hates 3's.
 
Yeah, I didn't come into the thread with the intent to dissaude anyone, just "temper your expectations" because I've had much more insane arguments from people who thought just because a RPi could run minecraft it could run windows and I'm like ... "no... it doesn't even work that way". I'm not angry.
 
People have some extremely misguided expectations and the point I want to drive home is that without an AAA developer explicitly developing for native Linux, Linux will continue to "not be a gaming platform", and a lot of this is self-inflicted because the Linux developers don't standardize on enough things to make it worth the investment.
 
The irony will be if Proton get's to a state where it can produce a self-contained runtime that works on any version of Linux, because then developers could just "support Linux" by including the runtime and not have to finagle with support of how to install their game on 400 different flavors of linux.

No, I'm being helpful by putting the spotlight on why Linux has drifted in the doldrums for being a general Desktop OS, never mind a Gaming platform. You may not want to believe it to be true. But it is.
 
If you want to see another example of "open source doesn't solve things", look at the SDL (simple directmedia layer) library. A cross-platform library that does all the annoying startup tasks to provide a general starting point to write a portable application.
 
What happened? They changed the API, they dropped support for things, and trying to update a game that uses SDL1 to 2, or 2 to 3 isn't straight forward. Instead of trying to keep the API consistant, they've instead ensured that anything built against SDL1, has no forward compatibility. You see this in emulators, but in particular, DOSBOX.
 
If at some point Microsoft breaks the "overlay" (which is the default) mode, ALL SDL games cease working without being able to reconfigure it, or recompile it.
 
 
I'm not telling you not to.
 
And relying on a single company to be in the Linux community's good graces forever is a disaster plan. Someone could buy Valve tomorrow and shut all of it down, or worse, the people who actually care about Valve existing could exit the company and sell it to Microsoft. Then what?
 
Short term thinking dooms everyone. We've enough evidence that something merely being open source doesn't save it from code rot.
 
And this is just the political argument a lot of Linux advocates push. "My computer, my software", and..? When all these vendors only write drivers for Windows, you are depending on a lot of good will from the Linux kernel developers to support your hardware.
 
 
I'm not. I'm saying don't expect parity, which is what unfortunately people have been pushing ever since WINE could run one game. People always demonstrate Linux boxes running emulators, not native games, and that is seriously disappointing and not encouraging developers to bring their games to Linux.
 
Been waiting for 25 years for there to be a good Linux that works at parity of Windows, or at Least MacOS. Still waiting. 
 
You know what promised us that? AmigaOS. Where is that? Pretty much dead. BeOS? Also dead. There is this obvious pattern that what we need as competition to Windows is "something is Windows in all but name, but not made by Microsoft" not "frankensteins monster that might, or might not run a Windows program and depends on the goodwill of the community to continue functioning and microsoft looking the other way.
 
Valve has a unique opportunity to make Linux at least somewhat usable as a gaming platform, but the previous 10 years, even though they supported Linux, they never helped bring any games TO linux that weren't their own first party games.
 
Proton will always be a bandaid, and because of how frequently Valve abandon's it's own projects to just let them waffle around, I woudn't trust Proton to not become code rot hell. It only works right now because of the Steamdeck, and if Valve decides the steamdeck isn't worth supporting in 3 years, they sure as heck aren't going to be putting more effort into Proton.
 

Have you seen global commitment to stop climate change? No. We just keep seeing the hurdles being moved.
 
Without negative growth in populations, and consequently negative growth in consumption there will be no reversal. It's not going to hit 1.5c and then plateau. It's just going to keep going up exponentially.
 

Relative mouse cursor position problems plague games that want to, or only operate full screen. This is already a problem in Windows when you force a full-screen game into windowed mode with DXWND. VM's complicate this.
 
The other side of this problem is virtual GPU's tend to ... just not work. You need to be able to access the actual GPU, and the best, and sometimes only way to do this is by letting the iGPU operate in the host OS and the dGPU in the client OS. You get around the relative window problems by simply plugging in a second mouse to use with the VM that only the client OS can see. So at that point you're basically just using the same keyboard and sound hardware from the host. 
 
If you are intending to play Windows games, on Linux, you should just play them on Windows unless the developer expressly blesses running it on Linux. This will keep you from getting banned in games using anti-cheat, and from getting blocked from broken DRM schemes.
 
I'm not saying "do not", but I am saying set your expectations MUCH lower for getting games to work on Linux. If it does not have a Linux binary, or isn't known to use a Linux native (eg Unity) game engine, then any mucking around you do with WINE or DXVK or Proton is as close as you're going to get.
 
On the other hand, if a game actually uses OpenGL or Vulkan, then even if it doesn't have a Linux native port, it might work under Proton/Wine with some coaxing. It's always going to be the DirectX and Windows native audio API's that are going to be an obstacle.

No. that is not true, never been true, and will never be true.
 
Linux always requires you to do more than "double click the icon" to get even simple "linux native" games to work, and you have to jump through lots of hoops to get WINE , DXVK, Proton and so forth to operate, and there will always be a few games DRM/Anti-Cheat that will prevent you from doing this altogether. 
 

No. that is not true, never been true, and will never be true.
 
Linux always requires you to do more than "double click the icon" to get even simple "linux native" games to work, and you have to jump through lots of hoops to get WINE , DXVK, Proton and so forth to operate, and there will always be a few games DRM/Anti-Cheat that will prevent you from doing this altogether. 
 

IPhones are stolen by street theives and organized crime, and then "refurbished" and sold to people who don't know the origin. All the stuff that is locked and doesn't allow activation, gets chopped up and sold as parts on eBay that amazingly some sellers have hundreds of.
 
You may not want to believe it, but this is pretty much what the majority of the used electronics on eBay and other market places. "Sold As-IS" = Stolen, "No Warranty", Broken "parts only", are all stolen devices if they are still under the manufacturer's warranty date. We know this is true because of the kinds of disputes raised at Paypal.
 
A three year old device, nobody is going to question if it's stolen or not. But a seemingly "new" device being sold for parts is.
 

Basically the PS4 Pro came out with the PS4 Slim, and then they discontinued the Pro and kept the slim for another 3 years. 
 
Personally, any time I've even bought a console, I waited for the "Slim" or whatever refined model comes out, because problems with the existing models. Like the Xbox 360 was notoriously bad, so I waited for a redesign , the S model. Which was fine, but then the hard drive died and I was like "well f microsoft then."
 
So I bought the Switch after the "OLED" model came out, but didn't buy the OLED model.
 

Probably so they don't have to change the board layout. Think about it. What they have right now:

If they change the CPU type, then they pretty much need to design a new PCB for it. If it has the same MB pinout then it can just be swapped straight over, and if the new GPU is just a die shrink of RDNA3/4 that fits in the same space of the RDNA2, then they don't need to even retool their process.
 
That said, I'm not sure why anyone would buy a "Pro" console. Does the existing console not play the games you want already? Did we learn nothing from Wii U? People don't buy incremental upgrades unless that is the only option. People who want this who already have a PS5 are going to be like "well I dunno, all my saves are married to that existing console."
 
 

It doesn't impress me. It creates maybe passable flavor-text/music you might hear in a shopping district inside a video game, but it doesn't sound real. It sounds like an AM Radio.
 
This is the problem with most "music" and "voice" AI, is that taken separately (See RVC) you can make a perfect clone of another song because all the AI does is "autotune" the voice B from the original voice A, but you use the same backing audio. All you've done is made a cover using the source voice, which to me isn't a "cover", it's akin to "nightcore"'ing a song where you just speed it up 50% and have done nothing else to it.
 
I'm not sure what the underlying process is for Udio because I've honestly just picked half a dozen different songs to check the genre adhere'dness but it seems like everything was washed with a noise filter that ranges from "radio" to "phonograph". I'd say most of these don't sound like what they claim to, and the ones that do, sound like they're inside a bathroom or hallway or down the street, or something.
 
They all lack "professional sound" mixing feel to it. If this was 1960, you could probably get away with it on LP or tape. 
 
Thinking about it for a minute, I believe I know what they did. They likely used commercial music as training data from different periods, because that would explain the incoherent levels of noise. The AI doesn't understand the "hiss" of a LP isn't part of the music.
 

Basically the PS4 Pro came out with the PS4 Slim, and then they discontinued the Pro and kept the slim for another 3 years. 
 
Personally, any time I've even bought a console, I waited for the "Slim" or whatever refined model comes out, because problems with the existing models. Like the Xbox 360 was notoriously bad, so I waited for a redesign , the S model. Which was fine, but then the hard drive died and I was like "well f microsoft then."
 
So I bought the Switch after the "OLED" model came out, but didn't buy the OLED model.
 

It doesn't impress me. It creates maybe passable flavor-text/music you might hear in a shopping district inside a video game, but it doesn't sound real. It sounds like an AM Radio.
 
This is the problem with most "music" and "voice" AI, is that taken separately (See RVC) you can make a perfect clone of another song because all the AI does is "autotune" the voice B from the original voice A, but you use the same backing audio. All you've done is made a cover using the source voice, which to me isn't a "cover", it's akin to "nightcore"'ing a song where you just speed it up 50% and have done nothing else to it.
 
I'm not sure what the underlying process is for Udio because I've honestly just picked half a dozen different songs to check the genre adhere'dness but it seems like everything was washed with a noise filter that ranges from "radio" to "phonograph". I'd say most of these don't sound like what they claim to, and the ones that do, sound like they're inside a bathroom or hallway or down the street, or something.
 
They all lack "professional sound" mixing feel to it. If this was 1960, you could probably get away with it on LP or tape. 
 
Thinking about it for a minute, I believe I know what they did. They likely used commercial music as training data from different periods, because that would explain the incoherent levels of noise. The AI doesn't understand the "hiss" of a LP isn't part of the music.
 

pfft. no.
 
It's always about preventing blackmarket chop-shops from proliferating.
The only reason it was possible to do this, was because some of the parts were "Chopped" must have been from stolen devices. Not just broken ones. If they were all broken devices, then he would have been unable to obtain the main SoC PCB, because they would have been "locked" from the previous user of the device, as they would have serial numbers that are tied to someone's iCloud account. Seriously if you think about it for 10 seconds, every part you find online on eBay that isn't a battery is from a stolen or broken device, but in order for someone to have hundreds of parts from a current model device, they have to be stolen. Nobody is buying the eWaste from bestbuy just to pick out the 5 year old iphones and send the rest to a shredder.
 
That's why companies should be required to buy back their old devices if they want to keep chop shops from operating. The company buys back the devices and then sends them to companies who remove the usable parts and are authorized to sell the parts, and the company will mark those devices in their inventory as "not stolen"
 
 

pfft. no.
 
It's always about preventing blackmarket chop-shops from proliferating.
The only reason it was possible to do this, was because some of the parts were "Chopped" must have been from stolen devices. Not just broken ones. If they were all broken devices, then he would have been unable to obtain the main SoC PCB, because they would have been "locked" from the previous user of the device, as they would have serial numbers that are tied to someone's iCloud account. Seriously if you think about it for 10 seconds, every part you find online on eBay that isn't a battery is from a stolen or broken device, but in order for someone to have hundreds of parts from a current model device, they have to be stolen. Nobody is buying the eWaste from bestbuy just to pick out the 5 year old iphones and send the rest to a shredder.
 
That's why companies should be required to buy back their old devices if they want to keep chop shops from operating. The company buys back the devices and then sends them to companies who remove the usable parts and are authorized to sell the parts, and the company will mark those devices in their inventory as "not stolen"
 
 

Summary
 A US state government has proposed a law requiring retroactively that generative AI models have their training data sources disclosed. 
 
Quotes
 
My thoughts
 This is nothing but good IMO. If we start requiring AI models to disclose what data they have ingested, we will have better quality models that can be checked against biases, and highlight which models are likely to result in output being lawsuit bait from purposely scraping/ripping commercial websites of non-free UGC material and other UGC sources. 
 
What I predict, is that if it does become law, commercial use of AI (eg ChatGPT) might slow down because the need to disclose will reveal which models have ingested copyrighted material should the output of an AI be claimed to be plagiarized of a copyrighted work. Can't use the defense of "well ChatGPT created it", when ChatGPT might have actually used the copyrighted work in it's training. Visual and Musical Artists will have a field day should it be revealed that their works were used to train a model and are being commercially used to replicate their styles.
 
What I don't see happening is any actual abandoning or shutdown of commercial generative AI use. They'll just change their TOS to put the liability on the end user for checking.
 
*UGC = User Generated Content, where the website is merely the publisher, not the owner. Think DeviantArt and Reddit.
 
Sources
 https://arstechnica.com/tech-policy/2024/04/us-lawmaker-proposes-a-public-database-of-all-ai-training-material/ 
https://schiff.house.gov/news/press-releases/rep-schiff-introduces-groundbreaking-bill-to-create-ai-transparency-between-creators-and-companies
https://schiff.house.gov/imo/media/doc/the_generative_ai_copyright_disclosure_act.pdf

Summary
 A US state government has proposed a law requiring retroactively that generative AI models have their training data sources disclosed. 
 
Quotes
 
My thoughts
 This is nothing but good IMO. If we start requiring AI models to disclose what data they have ingested, we will have better quality models that can be checked against biases, and highlight which models are likely to result in output being lawsuit bait from purposely scraping/ripping commercial websites of non-free UGC material and other UGC sources. 
 
What I predict, is that if it does become law, commercial use of AI (eg ChatGPT) might slow down because the need to disclose will reveal which models have ingested copyrighted material should the output of an AI be claimed to be plagiarized of a copyrighted work. Can't use the defense of "well ChatGPT created it", when ChatGPT might have actually used the copyrighted work in it's training. Visual and Musical Artists will have a field day should it be revealed that their works were used to train a model and are being commercially used to replicate their styles.
 
What I don't see happening is any actual abandoning or shutdown of commercial generative AI use. They'll just change their TOS to put the liability on the end user for checking.
 
*UGC = User Generated Content, where the website is merely the publisher, not the owner. Think DeviantArt and Reddit.
 
Sources
 https://arstechnica.com/tech-policy/2024/04/us-lawmaker-proposes-a-public-database-of-all-ai-training-material/ 
https://schiff.house.gov/news/press-releases/rep-schiff-introduces-groundbreaking-bill-to-create-ai-transparency-between-creators-and-companies
https://schiff.house.gov/imo/media/doc/the_generative_ai_copyright_disclosure_act.pdf

Summary
 A US state government has proposed a law requiring retroactively that generative AI models have their training data sources disclosed. 
 
Quotes
 
My thoughts
 This is nothing but good IMO. If we start requiring AI models to disclose what data they have ingested, we will have better quality models that can be checked against biases, and highlight which models are likely to result in output being lawsuit bait from purposely scraping/ripping commercial websites of non-free UGC material and other UGC sources. 
 
What I predict, is that if it does become law, commercial use of AI (eg ChatGPT) might slow down because the need to disclose will reveal which models have ingested copyrighted material should the output of an AI be claimed to be plagiarized of a copyrighted work. Can't use the defense of "well ChatGPT created it", when ChatGPT might have actually used the copyrighted work in it's training. Visual and Musical Artists will have a field day should it be revealed that their works were used to train a model and are being commercially used to replicate their styles.
 
What I don't see happening is any actual abandoning or shutdown of commercial generative AI use. They'll just change their TOS to put the liability on the end user for checking.
 
*UGC = User Generated Content, where the website is merely the publisher, not the owner. Think DeviantArt and Reddit.
 
Sources
 https://arstechnica.com/tech-policy/2024/04/us-lawmaker-proposes-a-public-database-of-all-ai-training-material/ 
https://schiff.house.gov/news/press-releases/rep-schiff-introduces-groundbreaking-bill-to-create-ai-transparency-between-creators-and-companies
https://schiff.house.gov/imo/media/doc/the_generative_ai_copyright_disclosure_act.pdf

Summary
 A US state government has proposed a law requiring retroactively that generative AI models have their training data sources disclosed. 
 
Quotes
 
My thoughts
 This is nothing but good IMO. If we start requiring AI models to disclose what data they have ingested, we will have better quality models that can be checked against biases, and highlight which models are likely to result in output being lawsuit bait from purposely scraping/ripping commercial websites of non-free UGC material and other UGC sources. 
 
What I predict, is that if it does become law, commercial use of AI (eg ChatGPT) might slow down because the need to disclose will reveal which models have ingested copyrighted material should the output of an AI be claimed to be plagiarized of a copyrighted work. Can't use the defense of "well ChatGPT created it", when ChatGPT might have actually used the copyrighted work in it's training. Visual and Musical Artists will have a field day should it be revealed that their works were used to train a model and are being commercially used to replicate their styles.
 
What I don't see happening is any actual abandoning or shutdown of commercial generative AI use. They'll just change their TOS to put the liability on the end user for checking.
 
*UGC = User Generated Content, where the website is merely the publisher, not the owner. Think DeviantArt and Reddit.
 
Sources
 https://arstechnica.com/tech-policy/2024/04/us-lawmaker-proposes-a-public-database-of-all-ai-training-material/ 
https://schiff.house.gov/news/press-releases/rep-schiff-introduces-groundbreaking-bill-to-create-ai-transparency-between-creators-and-companies
https://schiff.house.gov/imo/media/doc/the_generative_ai_copyright_disclosure_act.pdf

Doubtful.
 
The more likely scenario is that it would allow Nintendo/Sony/SquareEnix/etc to release "Nintendo switch online" types of services.
 
Like you have to understand that the most emulators are under licenses that would not allow them to be distributed on iOS because there's not a way to include the source code. Unless you compiled it yourself there was no way to legitimately put it on your device anyway. Not unless it was was a BSD/MIT license at least.
 
Certain games that you can buy on Steam right now, are emulators, or something close to it. For example:
This doesn't work. Like at all. Yet if you unpack the unity files, guess what's inside? Plain working ROMs that do work in any available SNES and Sega Mega Drive emulator. You just need to know which goes with what.
 
Another example, 
Sega delisted it, but it still exists if you bought it:
"SEGA Mega Drive and Genesis Classics"
Exactly the same as the Disney package above, at it is is an emulator, you can pull the games from it.
 
You can do this for dozens of other games, if you bought it on any original media, a lot of Square Enix games you buy on steam now, or previously on mobile, or on PSX, that were originally SNES, GB or GBA games, used the original game ROM, even though the "game" is little more than a script interpreter that understands the original binary. Like you would with ScummVM.
 
Does ScummVM count as an emulator? Does MAME count as an emulator. After all both of these are playing the original games without the original hardware.  But their source code license prohibits them being used on iOS, Playstation, Switch, Xbox, because they can't include the source code for the SDK for the game consoles, nor include the source code of the game.
 
Apple primarily didn't want emulators on their platform because of the possibility of using them to jailbreak the device and play pirated media. If a program has a fixed set of games that are licensed to be played (a la switch online) or is in fact a more generic emulator that works with a USB attachment to an actual "cart reader" that's permissible. Technically.
 
From a legal perspective, it would be smart for Nintendo to actually find a way to get Switch Online on the iPhone/iPad, because they could then say any other emulators of NES/SNES/GB are unauthorized and likely pirate software. But we all know they're not going to do this. I would expect Square Enix and SEGA to just release their games this way and forgo having to deal with both Nintendo and Apple's middle-grounds, and just deliver their old games directly to players.
 
One way or another.
 

Doubtful.
 
The more likely scenario is that it would allow Nintendo/Sony/SquareEnix/etc to release "Nintendo switch online" types of services.
 
Like you have to understand that the most emulators are under licenses that would not allow them to be distributed on iOS because there's not a way to include the source code. Unless you compiled it yourself there was no way to legitimately put it on your device anyway. Not unless it was was a BSD/MIT license at least.
 
Certain games that you can buy on Steam right now, are emulators, or something close to it. For example:
This doesn't work. Like at all. Yet if you unpack the unity files, guess what's inside? Plain working ROMs that do work in any available SNES and Sega Mega Drive emulator. You just need to know which goes with what.
 
Another example, 
Sega delisted it, but it still exists if you bought it:
"SEGA Mega Drive and Genesis Classics"
Exactly the same as the Disney package above, at it is is an emulator, you can pull the games from it.
 
You can do this for dozens of other games, if you bought it on any original media, a lot of Square Enix games you buy on steam now, or previously on mobile, or on PSX, that were originally SNES, GB or GBA games, used the original game ROM, even though the "game" is little more than a script interpreter that understands the original binary. Like you would with ScummVM.
 
Does ScummVM count as an emulator? Does MAME count as an emulator. After all both of these are playing the original games without the original hardware.  But their source code license prohibits them being used on iOS, Playstation, Switch, Xbox, because they can't include the source code for the SDK for the game consoles, nor include the source code of the game.
 
Apple primarily didn't want emulators on their platform because of the possibility of using them to jailbreak the device and play pirated media. If a program has a fixed set of games that are licensed to be played (a la switch online) or is in fact a more generic emulator that works with a USB attachment to an actual "cart reader" that's permissible. Technically.
 
From a legal perspective, it would be smart for Nintendo to actually find a way to get Switch Online on the iPhone/iPad, because they could then say any other emulators of NES/SNES/GB are unauthorized and likely pirate software. But we all know they're not going to do this. I would expect Square Enix and SEGA to just release their games this way and forgo having to deal with both Nintendo and Apple's middle-grounds, and just deliver their old games directly to players.
 
One way or another.
 

I'd probably not bother with the CPU. You can pull the GPU if you still have the packing box for it, otherwise it'll probably just get damaged.
 
Usually the primary rule for air travel is "buy an extra seat for extremely valuable equipment, and take everything in carryon" of which a seat might cost as much as the computer in some cases. Only check luggage that you can afford to replace.
 
Basically as others said, you want to take your data, on the SSD, and if you just swap a new computers SSD with the SSD you have, if it has the same CPU you can just pick up where you left off with most software not needing to be reauthorized.
 
If you're willing to spend the money to ship it, I'd evaluate how much you're paying to ship it versus how much it costs to buy it new, because generally people who travel frequently have to stick to mid-tier thicker laptops due to how much it costs, but also because ultrabook designs are easily damaged from handling at airports.
 
Like, electronics don't travel well at the best of times, even when things are sent commercially on pallets, they are often overpackaged and wrapped so they don't move and can withstand another pallet being placed on top. A single computer is likely to be damaged, even if well packaged simply from how it will be handled by the logistics company, FEDEX/UPS/DHL/etc literately drop stuff from planes which is typically an 8 ft drop.
 
I've seen computers packaged in wood crates, end up with inch-deep bends in the chassis from these drops. Hence, reiterate, avoid shipping the whole desktop.
 

Windows has had a variable EOL
 
If you include all versions of windows going back to 1.01, all versions of windows between 1.01 Thru to Windows 95 EOL'd in 2001
98/98SE and ME EOL 2006 ( 8, 7 and 6 years)
Windows NT 3.51 1995 to 2001 (6 years)
Windows NT 4.0 1996 to 2004 ( 8 years)
Windows 2000 2000 to 2010 ( 10 years)
Windows XP 2001 to 2014 ( 13 years)
Windows Vista 2007 to 2017 ( 10 years)
Windows 7 2009 to 2020 (11 years)
Windows 8 2012 to 2016 ( 4 years)
Windows 8.1 2013 to 2023 (10 years)
Windows 10 2015 to 2025 (10 years)
Windows 11 2021 to undefined
 
Now if you want to nitpick Windows 8, 8.1, 10 and 11 are the same OS. Microsoft basically rolled out sub versions starting with 8 that were equal to the service packs of NT 4. So each of these individual versions have small EOL dates. So Windows 10 1507 still has LTSB to 2024, even though GAC is 2017. 21H2 was released in 2021, and ends 2023, but LTSB is to 2032
 
So yes "Windows 10" has lasted 10 years, but each of the builds were 2 years.
 
Compare this with , say FreeBSD. Which 4.x had 7 years, 5.x had 5 years and every major version has been 5 years, but the point versions are frequently only 1 year. Which absolutely sucks except for the fact you can upgrade directly from 4.x to 13.x on the same hardware (personal experience) without ever having to reinstall things. So that's like having 25 years of support. Though it's not really that, FreeBSD has been removing hardware support and thus 13.x and 14.x will not run on the same hardware that was out with 4.x. FreeBSD also has immense code rot. So the ports packages are completely nuked when the OS goes EOL, so you are totally screwed the minute the OS goes EOL on that old hardware.. No upgrade path. Throw your server away.
 
I sometimes thought the cobol Y2K stuff was a bit silly, but here we are with the OS vendors deciding to arbitrarily nuke support for hardware just so they can release frequently. Linux is no better. Every piece of old hardware needs a maintainer for it's drivers or it risks disappearing there too.