The 2010s Were Very Successful For Wine Thanks To CodeWeavers + Valve's Steam Play

[Cybmax]

I kind of end up with the notion that "That is the way it is, no reason to change".

"How can we make wine better?" ** crickets **.
"Oh, i have this idea: " "Silence worm! You do not deserve to talk!"

The end.

[Weasel]

This has happened several times with "plain" wine aswell, and this is all dependant on the anti-cheat software in use. If a anti-cheat software is checking against know HASH of the .dll's or other file checks which anti-cheat software sometimes do, this WILL fail nevertheless.

And what happens on Windows if a Windows Update changes the hash of the dll and you play before any game update?

[Cybmax]

And what happens on Windows if a Windows Update changes the hash of the dll and you play before any game update?

Well, the anti-cheat detection does not work like an "instant thing" on your client when it comes to things like that tho. It works more or less like a "in the wild" kind of thing antivirus software uses, and in the case of an "unknown", you would have the client send this data to the game company. If suddenly 1 gazillion such reports started pouring in, would you think this would warrant an investigation? Yeah. THEN someone would check if something had caused this maybe on the driver side, or M$ side or whatnot. If the hash of this would come from an update, then this "unknown" would be flagged as "ok".

From a Linux point of view, the reports would be like.. 5. So, lets say a game company gets 5 "unknowns". What would be the best action? "Wham bam, ban hammer". Easy math.. its not ppl in the millions playing anti-cheat games using Wine/Proton, as most anticheat have serious issues under wine in general. (Like EAC or BattlEye).

And i am just using this as an example. It all depends on WHAT the anti-cheat checks, and how strict the company determines it is going to be ref. the case.

After several ban issues a couple of years back, Blizzard reverted bans for ppl playing using OverWatch. They ARE aware of ppl doing it, and state they do not provide support or any guarantee of success, but has proven they are willing to revert bans.

https://us.forums.blizzard.com/en/overwatch/t/can-we-get-banned-for-playing-on-linux/70929/3

https://www.altchar.com/game-news/blizzard-unbans-overwatch-players-who-used-linux-os-agF9y0G2gWjn

So, does this mean that the oh-so-perfect function of WineD3D would NOT have had any issues? Using WineD3D would be "undetected" in terms of anti-cheat?

I know you were not the one saying that, but please provide some evidence that when using WineD3D ppl cannot get detected or banned due to game anti-cheat.

Ppl DID get banned playing Blizzard games back in 2012 WAY before DXVK was invented... but ofc ppl would (just like religious ppl) claim that "Oh, nah, that was THEN, things are a LOT more perfect now!". And sure, it likely is. It does not mean it is useful to hide behind THAT as an excuse, cos if a ban happened with someone using WineD3D in 2020, it most certainly would be the game company's fault vs if it happened using DXVK, cos then it is "not using the correct wine". Don't hide behind ignorant claims.

https://www.cinemablend.com/games/Blizzard-Admits-Linux-User-Was-Wrongly-Banned-Offers-Refund-49339.html

So what happened here? Yeah, the game company decided it would listen. Does this happen with EA in the case of the linked BattleField V ban that was posted? Does anyone know EA to be a company that really listens to their customers? Anyone?
Nah, until you can actually show EA to have investigated this to be "illegal calls using the DXVK d3d11->Vulkan api", i choose to think EA just do not give a shit, and picks the easies way out "We have no idea what they are speaking of, but in case it is a REAL cheat, we choose to uphold the ban"... just like Blizzard DID at the start many years ago.

[oiaohm]

Well, the anti-cheat detection does not work like an "instant thing" on your client when it comes to things like that tho. It works more or less like a "in the wild" kind of thing antivirus software uses, and in the case of an "unknown", you would have the client send this data to the game company. If suddenly 1 gazillion such reports started pouring in, would you think this would warrant an investigation? Yeah. THEN someone would check if something had caused this maybe on the driver side, or M$ side or whatnot. If the hash of this would come from an update, then this "unknown" would be flagged as "ok"..

This not how it in fact works. Anti-cheat being like a anti-virus is true. If your incorrect behaviour matches some prior signatures incorrect behavour be it Linux, Windows, Mac, Console.... its instant ban always. its now can you appeal this.

So, does this mean that the oh-so-perfect function of WineD3D would NOT have had any issues? Using WineD3D would be "undetected" in terms of anti-cheat?

I never said WineD3D was perfect. The thing here is the conformance suite is critical to get bans over turned. Please note Blizzard linked developer added a test to the wine conformance test suite back in 2012. They did not tell us where in the game protection system this code was used and how but but they did provide what need to be fixed. The fact wine project was willing to fix the problem in a way that it would never happen again proves that it was not intentional on the users part so refund offers.

So you are referring the Blizzard case without understand what happened.

So what happened here? Yeah, the game company decided it would listen. Does this happen with EA in the case of the linked BattleField V ban that was posted? Does anyone know EA to be a company that really listens to their customers? Anyone?

Problem is the EA case is not the same as the Blizzard one in a big way.

Nah, until you can actually show EA to have investigated this to be "illegal calls using the DXVK d3d11->Vulkan api", i choose to think EA just do not give a shit, and picks the easies way out "We have no idea what they are speaking of, but in case it is a REAL cheat, we choose to uphold the ban"... just like Blizzard DID at the start many years ago.

No this is looking at the wrong party. Did DXVK add patches to improve Battlefield V without considering how the API/ABI behaved on windows yes. Do this is truly playing with fire Wined3d has this enough times by error without intentionally not considering the windows behavour .

https://github.com/doitsujin/dxvk

Manipulation of Direct3D libraries in multi-player games may be considered cheating and can get your account banned. This may also apply to single-player games with an embedded or dedicated multiplayer portion. Use at your own risk.

DXVK developer knows he doing behaviours that does not match windows and just put this warning on DXVK so he does not have to take any responsibility and dumps it all on the end user. So why did not DXVK users not expect to get banned and lose their money. Party because they had seen prior wined3d cases where at worse they would get refunds. This was also the users ignoring what the wine project does so that happens.

Really its been surprising that DXVK and been allowed for as long as it had without a major company getting upset. Please note EA developers and other game developers in past have also submitted to wine test suite. I would suspect DXVK has broken something in the wine test suite EA submitted as don't do this.

This is why the idea that you can test with real world applications only does not work either. We need means for maintainers of those real world applications to say this behaviour here the program might work but is not acceptable and is not windows behaviour.

We know of over 1000 tests in the wine test suite that DXVK fails with that wined3d does not with over 200 of those being submitted to wine from a developer associated with EA. There is every likely hood that EA has told the problem in the wine test suite. The issue is DXVK developer instance to test with real world applications and ignore the wine test suite. How many more times will people have to be banned before he wakes up this is not an option.

EA will not tell you the exact problem in exact game when it can be used as a cheat. EA, Bizzard and other companies adding to the wine conformance test they don't have to say its exactly this application and exactly here basically they don't have to point at the cheat. Wine test suite allows basically a mutual understanding between software provides and wine. Yes the wine test suite is a way of those software providers providing direction to wine without having to give away very much at all. This is why wine issues end users do end up getting refunds who stick to just wine provided source and binaries.

DXVK case where it clearly the developer does not care and has not attempted to setup understanding between software providing and them the result is simple end user gets marked cheating and loses their money..

Cybmax like it or not wine conformance test suite is not a optional thing. Wine project is very insisting that developers wanting to patch wine submit to conformance suite because this is a requirement to maintain the trust and respect from software providers to end users of wine. Without that trust and respect to wine end users they are going to get hurt financially.

Basically you care about your end users and you are doing something like wine you will make usage and development of the conformance test suite mandatory on your developers and allow developers from software providers to extend it.

[Cybmax]

This not how it in fact works. Anti-cheat being like a anti-virus is true. If your incorrect behaviour matches some prior signatures incorrect behavour be it Linux, Windows, Mac, Console.... its instant ban always. its now can you appeal this.

Yes, if the api call or whatever that DXVK uses is also the same as a KNOWN cheat uses, i can agree with that statement. This is however pure speculation, but the mechanic of "in the wild" type of virus or cheat does not work like this. It works by submitting "suspected behavior" and then some mechanic to evaluate this. It is not a instant kind of thing unless it is actually verified as being a cheat. Sometimes it is indistinguishable behavior between what a cheat does and what a new driver does. It is a lot easier to verify this behavior when it comes to a driver, and i will give you the fact that if a game company contacts eg. nVidia about a certain function, nVidia will provide the equivalent to wine conformance tests for them to verify and subsequently "unflag" this as cheating.

I never said WineD3D was perfect. The thing here is the conformance suite is critical to get bans over turned. Please note Blizzard linked developer added a test to the wine conformance test suite back in 2012. They did not tell us where in the game protection system this code was used and how but but they did provide what need to be fixed. The fact wine project was willing to fix the problem in a way that it would never happen again proves that it was not intentional on the users part so refund offers.

Until this happened to be a FACT, there WAS a lot of ppl actually banned for a long time. This was not a "fix was ready by next tuesday, and all bans were lifted" when looking back at it. It took a long time. Kudos to the Wine team by fixing this, but it is also not to say that will never happen again. And in the months it takes for me to get my ban lifted, i have "lost" my money.

No this is looking at the wrong party. Did DXVK add patches to improve Battlefield V without considering how the API/ABI behaved on windows yes. Do this is truly playing with fire Wined3d has this enough times by error without intentionally not considering the windows behavour .

https://github.com/doitsujin/dxvk

DXVK developer knows he doing behaviours that does not match windows and just put this warning on DXVK so he does not have to take any responsibility and dumps it all on the end user. So why did not DXVK users not expect to get banned and lose their money. Party because they had seen prior wined3d cases where at worse they would get refunds. This was also the users ignoring what the wine project does so that happens.

Really its been surprising that DXVK and been allowed for as long as it had without a major company getting upset. Please note EA developers and other game developers in past have also submitted to wine test suite. I would suspect DXVK has broken something in the wine test suite EA submitted as don't do this.

How many steam games have had proton users banned since its start due to proton using DXVK? I don't have the numbers in front of me, but for evidences sake i hope you consider Valve a "major company". It sure would be interesting reading, since i am pretty sure that DXVK does illegal api calls in Steam games too...

We know of over 1000 tests in the wine test suite that DXVK fails with that wined3d does not with over 200 of those being submitted to wine from a developer associated with EA. There is every likely hood that EA has told the problem in the wine test suite. The issue is DXVK developer instance to test with real world applications and ignore the wine test suite. How many more times will people have to be banned before he wakes up this is not an option.

Well, there are plenty of wine tests that fail when using anything other than a "Generic VGA (Microsoft)" driver too when testing on a windows machine, so i guess by that logic there should be a chance of getting banned if you use a proprietary driver from nVidia/AMD/Intel too perhaps? However, it COULD mean that the tests that companies like EA uses is checking for things wine have no idea about, since the conformance tests is basically reverse engineered "i think it works like this when looking at the description of the API call". (PS. Look at the bottom of my post for the question about this).

DXVK case where it clearly the developer does not care and has not attempted to setup understanding between software providing and them the result is simple end user gets marked cheating and loses their money..

Cybmax like it or not wine conformance test suite is not a optional thing. Wine project is very insisting that developers wanting to patch wine submit to conformance suite because this is a requirement to maintain the trust and respect from software providers to end users of wine. Without that trust and respect to wine end users they are going to get hurt financially.

Basically you care about your end users and you are doing something like wine you will make usage and development of the conformance test suite mandatory on your developers and allow developers from software providers to extend it.

Since this is the absolute truth, i am very much interested in learning about how this works. Can you please link to articles describing this mechanic in practice - eg. actual case files from EA or Blizzard where they describe what conformance tests needed to be done to verify this, and what changes needed to be done back from my previous examples of bans that got overturned AFTER wine provided the correct conformance tests? I am actually interested in this, and are not saying this to be sarcastic.

[oiaohm]

Yes, if the api call or whatever that DXVK uses is also the same as a KNOWN cheat uses, i can agree with that statement. This is however pure speculation, but the mechanic of "in the wild" type of virus or cheat does not work like this. It works by submitting "suspected behavior" and then some mechanic to evaluate this. It is not a instant kind of thing unless it is actually verified as being a cheat.

Except in the EA case DXVK is declared the cheat because it altering game behaviour in way that could give a user DXVK of an unfair advantage. Of course something like wined3d that is hindering games performance its not giving player an advantage so not a cheat so tolerated. Being better than windows is not exactly a good thing with games in a lot of cases.

nVidia will provide the equivalent to wine conformance tests for them to verify and subsequently "unflag" this as cheating.

Sometimes nVidia in these cases does what you say some cases they just provide note to game maker download wine test suite and note different reported here. Yes Wine conformance tests are not just used by wine. Conformance tests basically how you argue over if a ban is valid or not.

Until this happened to be a FACT, there WAS a lot of ppl actually banned for a long time. This was not a "fix was ready by next tuesday, and all bans were lifted" when looking back at it. It took a long time. Kudos to the Wine team by fixing this, but it is also not to say that will never happen again. And in the months it takes for me to get my ban lifted, i have "lost" my money.

You will see latter why wine project cannot promise it will never happen again.

How many steam games have had proton users banned since its start due to proton using DXVK? I don't have the numbers in front of me, but for evidences sake i hope you consider Valve a "major company". It sure would be interesting reading, since i am pretty sure that DXVK does illegal api calls in Steam games too.

Except Valve is a different matter. Valve with proton has developers in fact submitting patches to DXVK to change things. So Valve detects unfair play they patch DXVK. A few players have been caught out by instant connection drops with steam with they attempt to use old version of proton or host wine with different dxvk yet current proton works perfectly. About 10 percent of the total steam games or about 1/4 running in proton have this behaviour.

So yes Valve has the same kind of instant ban hammer just they don't use it as I nick your money and never let you play again instead they just absolutely drop your connection until you have a configuration that is acceptable and if you never get the game to run on your system give you a instant refund(yes this is Windows, Mac and Linux).

Out the game companies when it comes to dealing with cheats Valve is one of the nicer ones. So you get a game from valve with instant ban due to your graphics card they will refund as well. Valve is basically nice. EA or Activision Blizzard not so much.

Well, there are plenty of wine tests that fail when using anything other than a "Generic VGA (Microsoft)" driver too when testing on a windows machine, so i guess by that logic there should be a chance of getting banned if you use a proprietary driver from nVidia/AMD/Intel too perhaps?.

https://www.pcgamesn.com/amd-graphics-cards-pubg-ban

If you play enough different games Its not a chance you will get auto instant banned if you use proprietary drivers from nVidia/AMD/Intel it a absolute certainty at some point with windows you will get banned. Heck ban hammer land on people because they have installed like 2 copies of windows on the same machine on different drives for some reason as well due that randomisation wine test suite detects so you don't even need to change your hardware to have these problems.

Welcome to o great its not just wine that has the habit of end up on the wrong side of the game makers ban hammer. Conformance tests are required to dig your way out of this.. Basically playing closed source games on windows part of the experience is getting banned and since this is part of the experience wine cannot exactly fix this if wine in fact copies windows behaviour. Wine project all it can do as much as they can to avoid it but its not 100 percent avoid able. So wine test suite usage by developers cannot be allowed to be optional.

However, it COULD mean that the tests that companies like EA uses is checking for things wine have no idea about, since the conformance tests is basically reverse engineered "i think it works like this when looking at the description of the API call".

When optimising games for speed game companies do their own conformance/benchmark tests suites as well. Some part out those test suites by game companies end up submitted to wine. So something things wine test suite does not know about is true. DXVK case wine people know EA submitted tests for DX10 and DX11 are failing dxvk that pass on all Windows graphics drivers and wined3d so this is fairly much smoking gun.

Since this is the absolute truth, i am very much interested in learning about how this works. Can you please link to articles describing this mechanic in practice - eg. actual case files from EA or Blizzard where they describe what conformance tests needed to be done to verify this, and what changes needed to be done back from my previous examples of bans that got overturned AFTER wine provided the correct conformance tests? I am actually interested in this, and are not saying this to be sarcastic.

I have given you the basic over view of the process but there limit how much detail you can give before you cause trouble. Remember wine need working relationships with these vendors to get anywhere. So you will not normally see write up over any particular case exactly.

https://www.cinemablend.com/games/Bl...its-48696.html

Just like here were media says Blizzard was caught out in a lie because they said they did not test for Linux Exploits. The reality is inside these big game companies the left hand does not know what the right hand is doing. Its not a lie the person asked simply did not know. Wine project depends on this to resolve these bans lots of the time quickly. Like you will see at times a person in wine project dealing with person from a game company that does not officially support Linux or wine. Yes they will normally be approved by legal of the game company to be talking to the wine project and handing over test parts. But you cannot write an article about this process because they are like the right hand running like the security of the game servers. Not the marketing department or head of development.

Think about it this way wine project does up a write up where the left hand(marketing/development) at times works out that a right hand(server operations) is helping wine project and artical resulted in extra pressure demanding game on Linux that they don't want to-do that useful right hand end up not in the company any more by some means.

Yes some of these write up kill wine project contacts for a while. Lets just say the process sorting out incorrect bans is not nice and is very much like a back room deal with evidence and evidence is required due to a lot of distrust.

[duby229]

Except in the EA case DXVK is declared the cheat because it altering game behaviour in way that could give a user DXVK of an unfair advantage. Of course something like wined3d that is hindering games performance its not giving player an advantage so not a cheat so tolerated. Being better than windows is not exactly a good thing with games in a lot of cases.

i didn't even read your whole wall of crap yet, just a got the first paragraph and it is literally the dumbest shit evar!! So you really think wine is better -because- it hinders performance! Really??? Wow, just wow...

I don't think i can handle anything more stupid than that

[Cybmax]

I read the whole wall-of-utter-nonsense and i just.. wow.

Except in the EA case DXVK is declared the cheat because it altering game behaviour in way that could give a user DXVK of an unfair advantage. Of course something like wined3d that is hindering games performance its not giving player an advantage so not a cheat so tolerated. Being better than windows is not exactly a good thing with games in a lot of cases.

Really? DXVK is so much ahead of windows on EA games when it comes to performance, so that it gives such an advantage so its deemed a bannable offense? This just blows my mind! Honestly. So, if you get your hands on a couple of SLI 2080 "golden sample" that overclock like a mother... you probably will get hit by the banhammer then, cos it sure is an advantage with those 20 more fps! Yeah, that makes sense!

I do not know why i even comment on this, cos this is plain stupidity. I am sorry. That shit wont fly.

I am done. Sorry. The end. Nothing more to see here.

[oiaohm]

i didn't even read your whole wall of crap yet, just a got the first paragraph and it is literally the dumbest shit evar!! So you really think wine is better -because- it hinders performance! Really??? Wow, just wow...

I did not write it the best. But it is the reality the lower performing wined3d will be allowed to slide with minor behaviour errors. Higher performance DXVK will not be allowed to slide with the same errors. Closer you get to windows speed the more on point you runtime behaviour has to be if not users will get banned.

Really? DXVK is so much ahead of windows on EA games when it comes to performance, so that it gives such an advantage so its deemed a bannable offense? This just blows my mind! Honestly.

I did not write it exactly right.

Please note you don't have to have faster frame rate than windows. Frame rates being in what is classed as competitive for a lot of games over 60FPS is classed as competitive and once in that class the game developers anti-cheat will be looking for code path errors and will instant ban or disconnect(valve) on them.

Wined3d hindered performance means its normally outside the ban hammer range due to not in fact producing what is classed as competetive FPS as any code path errors the game does will not give you any playing advantage over another player so no ban hammer. This is how the ban systems horrible work.

So, if you get your hands on a couple of SLI 2080 "golden sample" that overclock like a mother... you probably will get hit by the banhammer then, cos it sure is an advantage with those 20 more fps! Yeah, that makes sense!

https://www.overclock.net/forum/78-pc-gaming/1731858-ea-banned-me-battlefield-v.html

Sorry to say this is EA we are talking about. They will drop the ban hammer on you for doing what you described is not probably will be hit by ban-hammer it that you will be hit by banhammer at some point. EA made games are not overclocking friendly. They are not the only game vendor like this.

You and I might not think this makes sense but its the reality of playing EA games and other vendors with rules like theirs.

Being better than windows is not exactly a good thing with games in a lot of cases.

I absolute stand by this statement. The reality is how touchy the different game vendor anti-cheats are. You make a system perform well better than the average. I should have wrote that more clear.

Being better than Windows with non overclocked parts is not exactly a good thing in lot of cases.<< This is what I should have wrote. This is a fact of playing games on windows. Then we have wine on top of that.

This is one of these horrible trade offs. If you are not exactly sure of how API/ABI should work be going after speed may not be the right thing.

Wine project is doing a lot to try to prevent users from getting banned. Slow development of dx12 support in wine is because wine developer are aware how touchy anti-cheat systems are. Some cases vkd3d has had extra debugging code put in to slow it down to where it was known not to be right so it would intentionally under perform to avoid the ban hammers.

duby229 you are really being dumb in a lot cases. Reality here is it not dumb to make under performing code when it results in user not been banned until you can get that section in fact working right. Welcome to the horrible trade off of working on wine graphics.


The reality of closed source gaming is a stack of stupidity that you have to work with to avoid end users getting banned. Yes some times giving people decent performance instead of poor performance is the wrong answer. Of course I wish the reality was that giving decent performance would never trigger game users to be banned but that is not the reality.