Performance is the issue. You'll be at a disadvantage thanks to virtualization. Youll have to sacrifice some frames. It's the same issue with gpu passthrough.

Yes, I was thinking of a similar concept with games run in containers. The issue is, that even if the container was inaccesible by the user, at some point the data needs to be processed by a CPU. Someone could catch the data there, but if they don't know how the game works at all, it will be harder

Google is implementing a similar idea in ChromeOS where Steam runs in a VM. It's called Borealis. It also relies on the Virgl and Venus drivers in Mesa for GPU acceleration.

@Watcher AI cheating is impossible to counter with local Anti-Cheat software anyway. The only way this method could be tackled is, if the Anti-Cheat also is AI-powered, but the performance hit would be higher on a client. Server side Anti-Cheat is also not that great since it's reliability heavily depends on the ping.

@@mritunjaymusale Manipulated game files is easy to detect. What's hard to detect are changes in the RAM or Cheats that hide themselves as an essential piece of hardware (e.g. a mouse input)

There are things out there that theoretically would work quite well. Using a virtual machine or even just a container for a game, if done right, provides a similar level of security. The key here is, that how the game works is obstructed from the cheaters. That means that they cannot don't receive location data for wallhacks, because they wouldn't know location data in an encrypted state. The only thing that these solutions won't prevent is AI, but maybe even an Anti-Cheat within a container would be possible, which can differentiate our wacky human accuracy from a bot. But it's hard to find a solution with encryption, decryption, the necessary APIs needed (GPU output, etc.) while performant and transparent. (What could the game transmitt from my system)

This is exactly why I don't want to game on Linux -- kernel level access. Let Microsoft do what they do well, keep their kernel and OS secure. If I had to dual boot to game, I'd take advantage of the full encryption option when installing Linux and Windows too if Ihad a Pro license. (Actually, I'd just spend the money to buy a Pro license and have Bit Locker. I'd grimace, but I'd do it.) Also, the two OS's would be installed on separate drives or series of drives. As it stands, it's just easier to have a dedicated gaming rig running Windows. So, I game on Windows and use Linux for everything else. Drive encryption, a two way firewall, never running on an administrator account -- maybe I'm paranoid. :-)

Yes, they can. You can format a HDD from windows too, so kernel drivers can do much more than normal software, so Yes, easily

I agree that for many things there should be hardcoded triggers. Like, ok maybe someone finds a way how to throw a nade across the whole map and kill someone, ok. But in CS2 it was possible to shoot them through like 8 walls with a sniper in some sort of burst mode. Like what?

dude, every comp game already has hardcoded anti-cheating software, the thing is, there are some sickos, who modify the input itself, not the game logic, like automatically moving your mouse down to prevent gun recoil, something a hard-coded anti cheat would never detect

You underestimate the complexity of modern cheat software.

​@@hunterzone4846 Probably, but there are many things in a lot of games which could be avoided by just coding the game right. For example: "wall hack". It could be avoided by just not sending the opponents to the client who are not visible. It's a more complex thing and needs a bit more server resource, but I think if they can make a game and maintain a lot of servers, they could easily handle this.

I hate to break it to you, but that's not how Anti-Cheats work. Can they collect data about you? Sure. Are they doing it? Some of them are collecting SOME data like your system configuration. But that's about it. Data collection is only worth doing, if you can sell that data to third parties, who can then proceed to make personalized ads, which is the most profitable way. AC-Vendors don't directly sell data, and this is known. They already make money by providing a service for companies, which as of itself is profitable (subscription model). Why would a company buy a subscription to a shady service? It would negatively impact everything if something were to get out. Think about this stuff in money terms, not in objective ones.

@@MichaelNROH I'd only truly believe that if they release the code for the anti cheat, which will never happen. I tend to lean on the conspiracy side of things these days 🤷‍♂️. Regardless I really like your channel and you make a good point.

Japan already outlawed cheats because it hurts the gacha game industry too much

@@MichaelNROH Data collection is a certainty especially considering that the companies that decide to use kernel level anti cheats have affiliations with the CCP (tencent).

It doesn't perform better? Not true I play Valorant and while I'm still pretty new around my 460 hours I have NEVER seen a cheater in Valorant

I'm a systems dev and ex-cheat developer and this is just false, the anticheat doesn't just """"trust the client""" and rely on ""security via obscurity"", it utilizes the server AND client to be able to gather enough data to determine if a user is innocent or cheating. Without the clientside anticheat, the server is effectively BLIND to non-obvious cheating, any good cheater can easily hide their cheats if there is nothing looking at the game's process. from the client-side. Most anticheats implement: - memory signature checks (compare patterns of memory within the game against a database of flagged cheat signatures on the AC server) - memory hooking checks (that pretty much any cheat needs to deploy in the game's memory) - foreign memory uploads (if there is a pattern of memory inside the game process that is not usually in the actual game, then upload it to anticheat server for inspection) - client-side heuristic checks (is there an overlay being displayed in the game that is not in the actual game)? - winAPI handle detection and blockade (for external cheats, they would be utilizing window's ReadProcessMemory and WriteProcessMemory functions) - DMA device detection (DMA devices can read the system's memory without actually doing anything in software) And the most important of all, pertaining the most to this context : Anomaly detection : if there is behavior in the kernel or anticheat daemon/driver that is not typically how it should be behaving, the player is forcefully kicked out of the server by the anticheat, this is not foolproof, however it is a strong mitigation for direct attacks on either the anticheat daemon/driver, or modifications of the kernel.

@@ChristopherGray00 You're contradicting yourself. "Without the clientside anticheat, the server is effectively BLIND to non-obvious cheating" and yet you say "the anticheat doesn't just trust the client". If your server is blind to a client that lies to your server about the data it has gathered then you are TRUSTING THE CLIENT. Memory signature checks? Sure boss, here's what you expected to see. Memory hooking checks? Sure boss, here's what you were expecting. Foreign memory uploads? Sure boss, here's what you expected to see. Client-side heuristics? Here you go boss, this is the heuristics you expect to see from gamers who aren't cheating. And on and on of course. Anything is open source if you can read the circuit board or otherwise reverse engineer, but security by obscurity is just a bet that the attacker won't go that far because its just a game so surely they won't put in that much work.

@@Nerd2Ninja The client checking to see what is going on in the game's memory is not trusting the client, you seem absolutely clueless as to the topic at hand, bypassing one mechanism within the clientside AC does not then magically bypass another mechanism within the serverside AC. If i write a cheat, in which the clientside AC has no signatures for, and can insta head lock aim in a randomized fashion and kill everyone on the team instantly, i didn't bypass the anticheat, the server can still see that. Were you intoxicated when typing such a stupid comment? Actually curious

Wine can't natively run drivers on its own, however there is nothing stopping a program from interfacing with wine, or a modification of wine that communicates to and from a driver.

This will only shift the root of trust to ring -1 though, which opens another round of cat and mouse game just like when they shifted from ring 3 to 0.

@D Reaper exactly, too much effort for a shittier counterstrike. I got my mates ro play cuned 2 instea dbc if the mac user in the office and our friendly neighborhood toaster gamer can play it at 60fps

Sure, as long as they stay exlusively in their own lobbies. You are right when you say that not everyone wants to grind a game, but that's no excuse for giving yourself an edge over those who want. Competitiveness needs fairplay or simple fun, even when loosing, automatically turns to frustration.

@@MichaelNROH You're hitting the nail on the head here. Competitive gaming presents a unique challenge when it comes to addressing cheating. Some players prioritize efficiency above all else, often at the expense of fair play. It's a reality that competitive gaming grapples with, where the term "competitive" can sometimes be synonymous with exploiting any available advantage. Trying to dissuade players from cheating in such an environment is no easy feat. For many, time is a precious commodity, and spending hours honing skills in a game can feel like an inefficient use of it. The stress and frustration that come with intense training can often overshadow the enjoyment of the game itself. That said, promoting a balanced approach is crucial. Instead of outright condemning aggressive cheating, advocating for cheats that align with the standards of professional gameplay, such as AI aim assist programs, could potentially create a more level playing field. These tools offer a way for players to enhance their performance without resorting to blatantly unfair tactics. By encouraging the use of cheats that emulate the behavior of professional players, we can foster a more respectful and enjoyable competitive environment for everyone involved. It's about finding a middle ground where players can still pursue excellence without compromising the integrity of the game.

I think it's also about the potential. I don't have any numbers because no one is releasing stats, but I don't think that from the actual playerbase, only a fraction is cheating. Those of course stand out from the crowd, because even in a 10 player lobby, there aren't that many cheaters (cheater per game). I think they are scared that the small percentage of Linux Gamers becomes a cheater only plattform because they switch. Stupid approach, but it's probably what's going on

The performance cost and heavy dependency on good pings makes this very hard to achieve. The netcode needs to be perfect to sync up if you could see an enemy that was moving behind a wall for example

@@MichaelNROH i see , thanks so much for answering, we hope this get's a fix sooner or later.

@@MichaelNROH I do feel like it's possible to mitigate this (somewhat) by doing movement prediction, as we've done on clients for decades now. But there'd still be a bunch of issues with this.

Battleye has userspace-support on Linux, but Vanguard has a huge problem. It ONLY runs on Kernel-Level. There is basically a whole rework needed, and frankly the potential player numbers on Linux are not that high for that.

@@MichaelNROH Yes but I am waiting for company named Ubisoft to enable proton support flip one f**king button and let us play games like Rainbow Six Siege! What is so hard in it, game even support Vulkan out of the box and plays really well.

AI Anti-Cheats probably can't ever be prevented unless you have an AI that counters them. There are some differences in how consistent a human behaves. Nonetheless, the problem is still the identifying.

@D Reaper easier said then done when you enjoy playing games. I don't like this "don't support" or "don't play" mentality. Just cause a small minority doesn't play them doesn't mean they will change. Like I said things need to change and hopefully they do soon before all the momentum the steam deck has built wears off.

I slightly disagree with this. Companies don't really have the power to choose platforms if they want their games to become a blockbuster. They need to choose platforms, that the MAJORITY of potential players uses. If a Steam Deck would sell millions of units and people start playing on that more, then they are forced to take action or otherwise they would loose players over time. They also have some guruantees to fulfill. If it was up to Bungie, then they would release Destiny on every single platform they could to increase the playerbase and margin. However, if that results in an explosion (or even potentially), then they will certainly loose great numbers of players. Don't hate the company for trying, hate the Cheaters.

​@Ransacked it's a domino effect. The more people that stop playing the more people will continue to stop playing. These companies will Make changes to please an extra 10-15%

Open Sourcing Anti-Cheats is not being done because it is less a security issue, but more a marketing issue. People make money with providing Anti-Cheat solutions and open sourcing that would lead to several issues. Most of all the exposure of methods which could lead to a temporary increase of cheating exploits, while the community is not allowed to fix issues themselves. (Server validation and licenses) Open Source Anti-Cheats could only work if: * The community is allowed to contribute * It offers commercial support for companies (Warranty and all that) * It somehow needs to be verified, that it is an official release which was not edited

​@@MichaelNROH honestly all of those things seem decently doable imo. The first one is just "the people behind it need to understand open source isn't just a buzzword", the second one could be a ZorinOS situation where there IS an organization/company behind the project that can offer support, but the actual thing itself is completely open. (yes Zorin pro does have some extra features, but let's be real the main reason people buy it aside from a fancy donation is for the support) The third one, well there I can't say anything for certainty unfortunately. I'm not a cryptography expert, I understand the basics of hashing, encryption, pub-private key pairs, etc. buuuuut definite statements on what is and isn't possible is a bit out of my purview. I'd imagine it probably is possible to verify it with some black magic cryptographic wizardry, but I wouldn't know how to say anything for certainty. In THEORY I could see a system where the internal variables are isolated in memory, and then the rest of the memory itself is used as an encryption key or something, (thereby requiring people be running an entire dummy instance of the entire anti-cheat in order to be spoofed) but, again, I can't speak with certainty. All I can say with high confidence is that I've seen good cryptography do some seriously absurd crap, "zero-trust computing" just seems completely impossible to me but, apparently it's a thing. Given that, I find it hard to believe you COULDN'T create some way of verifying it's correct, buuut the exact implementation, how wide reaching it is, how effective it is, etc. is beyond me. The only real point I'd disagree with is the aspect of marketing. In regards to licensing, if it did follow a "support" business model, so companies could optionally license support to fund the organization that manages the product, rather than a central company which wholly develops the product, you wouldn't really need have a dead period where the community can't contribute for licensing reasons. Additionally, since you're basically outsourcing most of the development and your mainly just guiding the project, you could offer much lower prices to license it, so smaller games would pick it up as a nice drop-in way to get some anticheat of anysort without breaking the bank. Then, once it has that userbase, it could grow up from there as people realize this low-cost solution is equally, or more, effective. I suppose I should clarify that my point is basically that "an open source, community driven (and potentially organization or company managed) anti-cheat would have every capacity to work equally as good, or superiour to, current anti-cheats". Obviously existing anti-cheat companies are going to have a problem with their business model being ripped out from under them, but I'm looking at it specifically from the community standpoint here. It would take some serious organizational work, but I can't imagine it being impossible. With some elements of corporate DRM already being baked into the kernel (unfortunately mind you, seriously, DRM being unnecessarily baked into the kernel is just a no-go for me when it patently doesn't work, just about everything ever can be found on 1337 so DRM only gets in the way of normal users) and proprietary blobs being common, a completely open source, user-first anti-cheat seems entirely reasonable to me. Maybe some hardcore FOSS guys with hardcore Stallman fanfic tatted on their junk would take some serious issue to it but I think for most people, if they know it isn't spying on them, if they know it isn't hurting their performance, and if it lets them play games, I don't think most people would care. Sure, "people don't care" isn't a complete justification for something (people don't care about a lot of things they should) but generally, so long as I still know it's honest and has my best interest in mind, I don't really see any reason to hate something just because it's goal is to limit what I can do. (at least not if I voluntarily install it that is, or have some way of removing it if I don't want it) Maybe I'm wrong on that last point though but thats just how it seems to me.

Maybe try to think about a "community anti cheat system". Something like crowdsec who collect IP address who try to hack into server and share those IP with all the community. This could help to identify people who are cheating and blocking them everywhere... Then the cost to cheat could start to be very expensive... Enforce static IP everywhere that is possible so people can't change their own IP to easly... When people will start to be banned for few hours, then few days then a full month on every gaming plateform because they have try to cheat on a game, then they will think twice about using such tools... It's impossible to fight against hackers, the only solution is to always increase the price to use such solution and in the same time, reduce the cost to fight them... I clearly think that there are some idea to look into...

@@LtSich people have tried crowdsourced unified ban lists before, they've failed miserably every time. What you would need is a FOSS detection engine that can hook into a per-game banning system.

@@robonator2945 This is exactly what crowdsec is actually doing.... A french company... Not used for game atm, but as long as you develop something to detect cheater and report those IP to them then you should be able to do the same for gaming... I use that to protect my server (again, no gaming atm) and it's really good. I'm sure that if the dev and this company work together they can build something similar for gaming...

That's because "theoretical", the Windows Kernel should be inaccessible to Cheat-Makers and Microsoft takes responsibility for that. On Linux, you can't really blame anyone, if someone just recompiles the Kernel. So the question is, why keep maintaining Kernel-Support for a small portion of gamers, when almost everything is open to Cheaters? With growing gaming numbers, Kernel support will follow, however it probably won't be much more effective against "Know-Hows".

The problem is not the compatibility most of the time, but just the Anti-Cheats. A VM won't resolve this since many ACs detect virtual machines and ban your account because of that.

Yeah it takes alot of effort to trick your vm into thinking its not a vm almost to much effort honestly im pretty sure compiling gento 3 times would be easier then doing this dualbooting would be way easier but honestly i really dont want to have to run closed source anti cheat witch has full accses to my cumputer from a coropration im just really concious about my kernal and whats going in there alot of malware exploit kernal anti cheats and bypass firewalls running it in a virutal mechine is more secure but most of the time if all your doing is playing video games you probbaly wont get hacked i gues im just really paranoid about my kernel@@MichaelNROH

Well yeah, but also no. Many games have those intrusive Anti-Cheats because they are competitive. And frankly, even on a fiber connection, the input, output latency is way too high for that

@@MichaelNROH my idea is more along the lines that player pc doesn't know where other players are in the game. If you can't see the other player ( being behind a hill ) you should not have info they are there at all. Biggest problem with cloud is bandwidth, compression and decompression. If server is at the same place latency stays the same for each packet. Maybe it could be approached with local render.

Yes and I think it's entirely possible to iron out most of the stability/latency issues. In fact, there's been some interesting work going on here in the direction of AI, allowing for insanely good real-time streamed video compression. Just currently requires you to have a 6000 dollar graphics card in both the server and the client lol.

Cloud gaming has it's own issues and can only be used by those with fantastic Internet connections

@@NightRaven5568 Maybe in a 100 years 1 gbit will be like 16 mbit today... lol But I agree.

It's not impossible to maintain since the Graphic Card drivers are on Ring 0. It's all about the execution

@@MichaelNROH Ring 0 Device drivers like Graphics ship with the Linux OS, Sometime some OEM Linux Distributors disallow the final user to load Kernel Modules, meaning you cant load your modules nor anyone after the Linux Kernel was shipped

@@MichaelNROH Some OEM Devs Turn off the insmod command, making it very hard to load Kernel Modules

​@@MichaelNROH the point is... unlike Windows, Linux does not have a stable kernel ABI. That means that without compiling the kernel driver SPECIFICALLY for the kernel version used it simply **will not work**. Also, loading these "out-of-tree" modules taints the kernel which makes its logs completely and utterly useless for bug reporting kernel/driver issues etc.

It's a technical challenge, but feasible

I think the reason on why Valve hasn't released Steam OS 3 for regular PCs is, that they have troubles with certain Hardware. An example of this would be Gamscope and NVIDIA, which doesn't play nicely.

That sounds expensive though 😅 I guess, what's really happening is, that they watch several games simultaniously

They don't need to install it. There is a difference between not being able to play a game and not wanting to play a game

I really like your statements, even though I find it very interesting how people "slip" into cheating. I myself somewhat come from the opposite side, whereas if I were to start cheating, I would just embarrass myself on local coop. Plus, maybe playing mainly on a console further set that mindset. The second point is kind of hard. Enforced by law does not mean punishements like prison or something like that. I mean, this is technically breaking a contract that they signed and they can be made responsible for losses if players stop playing after facing them. This is something that can happen at any point, since Cheaters signed the contract that they won't cheat and revenue loss lawsuits do exist. Can get very expensive, but the company decides if they want to sue or not. It's not controlled by the state or country is what I'm saying

@@MichaelNROH From what i observe from some of my friends, rather than get embarrassed, they just admit it they use cheat when asked and move on. When it comes to coop or partying up, what quite embedded in their brain is "My friend is platinum, and i don't have time to get there in time when he ask me to party up next week, so lets get it on", These kind of cheaters usually not cheat when you play party on competitive type of game (ranked, mmr, etc), and only cheat when play solo. Looking at the the paragraph above, "addiction" might be too much, but for cheater in my example, cheating becomes something "hardwired" to the brain. When you want something from the game, most player think if they have the time, should they clear some schedule, ask to be carried, but cheating is kinda a "sins", while for the cheater in my example, cheating just a method they can just use as if it is something normal (even though they personally know it is wrong). The moral brakes when they consider to cheat just not there compared to normal player.

bruh

Well that mindset was already there. What is missing nowadays is the social factor when playing video games. Back when LAN Parties and local Splitscreen was a thing, the mindset was different, since people could get to you in real life. We weren't trying that hard. With this face-to-face border that is the internet, a lot of people feel like they are more unhinged since they often only play in small groups together, instead of against each other. Since no one could hurt them, they use cheats more increasingly. Btw. Server side AC is a thing, however the processing power and accuracy for many players at the same time, is not a good for finances, but also the player experience. This only works for some Games. Valorant or CS:GO with 128 tick and 10 concurrent players is a nightmare for a server to calculate.

@@MichaelNROH Yeah, we all know their excuses, but as a programmer I know they're all false. It doesn't take up dramatically more processing power to do the kind of processing necessary to prevent cheats. Sure, they won't defeat "AI" cheaters, but they can't do that anyway. The majority of players won't have that capability either, for now. As for the mentality, anonymity plays into that a bit, but once someone gets to be a top player they generally have their name and face revealed. I don't know if anyone will get what I mean, but I think it goes deeper than that. I think it's all part of the moral decay of our civilization that encourages and forgives bad behavior. They look at the situation as black and white and have taken us from what they viewed as an extreme to what is actually an extreme. For the games industry, I think the only sensible way forward would be to open source everything. Make buying a game the same as buying a book or a piece of music. You pay the artists for their creation, but you get to experience it wherever and however you want. I know you don't agree, but only because you haven't seen how it could work. Game engines these days are mostly incrementally updated, and not everyone uses a close source engine anyhow. The bulk of the money taken in by a game should go to the artists since they do the bulk of the work. Ideally, the publishers would get a lower cut in this scenario.

Nahh, not really. I believe it just took some time for cheat makers to advertise themselves (money) and due to the growth of streaming services, more people feel bad if they can't keep up with the pros. The growth of the internet is essentially the growth of cheat sellers and consumers as well.

@@MichaelNROH Games being paid for and popular / good servers being well moderated meant cheaters were quickly taken care of -- they still existed but could be banned quickly, so it wasn't much fun to cheat. Because these companies now monetise in-game, and because there are monetary incentives to cheat, they need to do their own moderation and the problem is now much worse. I think anti-cheat is just DRM with reputation laundering built in. Thanks for replying to my comment, I'm a recent subscriber and your videos are excellent.

Easy: Old games were simpler. Also, people expect more and more from video games, but the number of developers working on a game can't be scaled infinitely, which causes more time pressure and more mistakes to happen.

Still would not work since you would need to encapsule everything related to the game. Audio, GPU, Mouse, Keyboard, etc. Every open driver could potentially be an entry point for cheats, and that is why a seperate Kernel module could only mitigate the issue, not get rid of it

Running Windows in a VM is not the way to go, espececially when the problem is related to Anti-Cheat. Using VM's is bannable if detected

Sure, but that still won't prevent people from easily disguising cheats as drivers

@@MichaelNROH and it would stop those as well as it does on windows because it has parity access level to any spoofed driver modules or w/e. Basically then we will be a feature parity with windows and able to be supported like windows.

I'm really not sure about if VAC actually works or if it is even improved upon. From all of the people that got banned iN my CS:GO matches (forgot where to look that up), I believe that the majority got banned with their Overwatch system (People reviewing the game), not VAC itself. A ban also needs to happen ingame and not after a while.

@@MichaelNROH yeah lot of people get banned by overwatch system. Vac isn't very effective at banning.

@@MichaelNROH i think they may have started using some new anticheat system in csgo cus i hear there are lot of bans recently and it doesn't seem to be just a normal vac wave. I think they may have automated the process of adding new cheat to vac so it will ban those cheats if it sees it cus previously i think they had to add those manually.

It wouldn't really do anything as they would need to still open source everything besides their proprietary Anti Cheat "blackbox". Same results

Legally speaking, emulation without owning the game is illegal though (ROMs from the internet). Not sure about owning the console as well.

@@MichaelNROH Real speaking, most emulation is for games so old fishing them on the 7 seas doesn't hurt anyone, not even the companies that created them. And in that case not even a moral issue, Imo. And I feel like doing that with games that DOES hurt the companies is still a MUCH smaller problem than this anti-cheat (specially since it may actually be morally fine to sanction DRM crap) But that might just be because of the Linux bubble I'm in lol.

Some server side Anti-Cheats do exist, however they face a lot of problems. For once the processing power and therefore the cost of running several game instances becomes more expensive. Another thing is latency. If you shoot at an enemy, the server has to calculate the chances of being legit, and this takes time. On the client's side it's much faster since everything got processed already.

I am SURE this is possible but it's just one of those things that companies won't pour in any money researching for because it's just cheaper to keep going with the current system.

I won't go back to Windows until I absolutely have too. If I play games on Windows, then telemetry data is telling game devs that I prefer to play on Windows. That's why I will continue to play everything on Linux

That's because most people use Windows. If Linux had the same amount of people and Windows was the opposite, then the story would look quite different

@@MichaelNROH i mean... thats the point? i mean, sure nexus mods will become simpler to use over time on linux, but who's to say that anticheat wont get better as well as time goes on?

Server side Anti-Cheats are not ready for deployment. Such servers need more processing power, but also guruantee a good latency. A game like Valorant or CS:GO with a higher tick rate (64 - 128) should not feel like a 16 tick game because the Anti-Cheat takes too long to calculate 10 players at once.

@@MichaelNROH But they also shouldn't be inaccessible to people who don't want to run Windows... meh. I know this argument sucks. But still lol

Because it's the best copright free thing I could find 😅