There are no images here.

Have you tried PlayOnLinux and juggling various versions of Wine? I often find the ability to change from one Wine to another on the fly a boon.

When I tried it with Wine with a GTX 650 it was running fine.
I didn't play much but the I never encountered any issue on the main menu, benchmark and during a short gameplay.

The only rendering problem that I had was with the Nouveau driver when using Wine Gallium Nine but it was different than yours (shadows not correctly render).

Just FYI Wine runs a simpler version of the game using DX9 which isn't the same as the more complex graphics path DX10/11 that the native version uses so it's not a similar comparison.

Also the Native version doesn't have all these glitches ;)

No Eddie. This is about FPS performance comparison at the same Quality presets (normal) and same hardware and not cosmetic details imperceptible at first glance... :)

The fact that I have more FPS performance with the windows version of the game running via Crossover 15.2 than the "native" Linux version with the same GPU, the GTX 650, is a point for Windows and Codeweavers, not for You...

If I need a much more powerfull hardware for to play a Linux port of whatever game than the hardware needed for to play the Windows version of the same game running on Windows (or Linux via winE), is a bad thing for the Linux gaming... A very bad thing if that game is an old shovelware game.

I don't know what other linux users think, but I want NEW GAMES, from the current year (or, at least, last year)...
I don't like to wait THREE years (or more) for to play a game on my beloved plattform...

Bad ports of older shovelware games...
good ports of older shovelware games...
the lack of NEW games for Linux (well or bad ported)...
cancelled ports (like Batman Arkham Knight)...
games available for Linux only on Steam but DRMFREE on windows (like Metro Redux and Dying Light)...
and that good idea that is OriginAccess (Valve should take a note), are the things that made me consider to have a Windows 7 PC just for gaming...

In less than a month I will finish the assembly of my Windows 7 machine (I don't like dual boot, I prefer to use two Pc's with a KVM switch) and I will compare the native windows version against your "native" linux version... Let's see who is the king of FPS performance.

I don't like to use Windows, but if I'm gonna use it, I don't like the idea of to "pirate" the OS like most people do, that's why I purchased three Legit Windows 7 Oem licenses. They cost me about 200U$D each one.. (I didn't purchase the three Windows 7 oem discs at the same time, by the way.. I'm not rich)

I will not leave the use of my Linux distro... I will keep a lot of my gaming on Linux... I will keep buying games for my [new Linux Steam account](http://steamcommunity.com/id/DigitalCHE_Reloaded) and not for [my old windows steam account](http://steamcommunity.com/id/Digital_CHE), like I did with Dead Island Definitive, Shadow of Mordor, ARK: Survival Evolved, SOMA, Shadwen and Alien isolation.

This seems silly. You're saying on one hand the wine version runs better, but on the other hand it doesn't run properly. Eddie turns around and makes a very valid point, the native version doesn't suffer from the same problems. His point is just as valid as yours. Saying Wine did a great job because their software runs the game faster but with artifacts all over the place and at a lower dx equivalent but Feral did a bad job because it runs slower at a higher dx equivalent and has no artifacts is just dribble. If you want to run Windows because the game performs better that's cool, go ahead. Some people don't want to, and they don't want to run it Wine and have it not work right or have to mess around with different Wine settings and versions. It isn't a good comparison, the dx9 version at high settings isn't the same as high settings on a dx10 or dx11 level. Maybe they could have made the game faster using dx9 equivalent code instead, then some people would knock it because it was using outdated graphics. It isn't a bad port, but it isn't a perfect port. You think they can do better and want to say so go ahead, but please don't bash one of the few companies doing good work for us because they couldn't do better than the Windows version with a fraction of the time and money, with code they didn't create in the first place.

In case you weren't aware, they can't just port any game they want. The game owners decide what they want ported and what they don't. I'm pretty sure WB pulled Batman because of how badly it did on Windows, nothing to do with Feral. Feral did bring X-COM 2 day one.

I kind of get where you're going with this, but I don't agree with your conclusions. I don't think Feral is bad for Linux gaming at all, I think they are one of the best things we have going for us.

Quoting: m2mg2This seems silly. You're saying on one hand the wine version runs better, but on the other hand it doesn't run properly.

You got it wrong.. I said the graphic glitches of the windows version running via wine are present only in the Maxwell Gpu's family... In the kepler GPU's everything works fine... That's why I think is a driver issue.

Take a look of the windows version running via Crossover with a GTX 650:

Quoting: m2mg2It isn't a bad port, but it isn't a perfect port. You think they can do better and want to say so go ahead, but please don't bash one of the few companies doing good work for us because they couldn't do better than the Windows version with a fraction of the time and money, with code they didn't create in the first place.

In case you weren't aware, they can't just port any game they want. The game owners decide what they want ported and what they don't. I'm pretty sure WB pulled Batman because of how badly it did on Windows, nothing to do with Feral. Feral did bring X-COM 2 day one.

I kind of get where you're going with this, but I don't agree with your conclusions. I don't think Feral is bad for Linux gaming at all, I think they are one of the best things we have going for us.

About Batman. The fact that the windows port is a piece of tech crap, doesn't mean that the Linux and Mac ports will be the same crap.. Those ports can have a better performance than the Windows one if the porters do the port from the original P$4 version and not from the broken windows port...
A port from a port is not a good idea.

I think We have a different concept about game porting.
For me, porting a game to Linux is to take a game that was made previously for another plattform, such as Windows or P$4, and remake it using another game engine 100% native of the target OS (in this case, a native OpenGL linux engine), but with the same graphic assets, same sounds, same game design, same gameplay, same maps, same cinematics, etc.. The final user must not notice any audiovisual difference between the original and the Linux remake... Or yes, The 3D models and textures can be modified for to look better than the original, especially if the original game is an oldie...

If You want to have an idea of what im talking about, Remember the kind of "homemade port" of Bioshock from UE3 to Cryengine3? ... Something like that..

If, someday, I have a game development studio / porting studio, I will use this technique.. Maybe it will take long, but the final performance will be better.

As far as I know, there are two kind of porting contracts:

1:The original publisher of whatever game hires an studio for to make a Linux port...

2:An studio / publisher wannabe acquires the publishing rights for Linux of whatever game from the original publisher.

I don't know why, but linux porters like Feral and Aspyr refuse to clarify the situation.
Do they have the publishing rights for Linux of the games they ported, like the second option, or they are just contracted companies, like the first option?

Quoting: Comandante ÑoñardoYou got it wrong.. I said the graphic glitches of the windows version running via wine are present only in the Maxwell Gpu's family... In the kepler GPU's everything works fine... That's why I think is a driver issue.

Take a look of the windows version running via Crossover with a GTX 650:

External Media: You need to be logged in to view this.

External Media: You need to be logged in to view this.

External Media: You need to be logged in to view this.

About Batman. The fact that the windows port is a piece of tech crap, doesn't mean that the Linux and Mac ports will be the same crap.. Those ports can have a better performance than the Windows one if the porters do the port from the original P$4 version and not from the broken windows port...
A port from a port is not a good idea.

I think We have a different concept about game porting.
For me, porting a game to Linux is to take a game that was made previously for another plattform, such as Windows or P$4, and remake it using another game engine 100% native of the target OS (in this case, a native OpenGL linux engine), but with the same graphic assets, same sounds, same game design, same gameplay, same maps, same cinematics, etc.. The final user must not notice any audiovisual difference between the original and the Linux remake... Or yes, The 3D models and textures can be modified for to look better than the original, especially if the original game is an oldie...

If You want to have an idea of what im talking about, Remember the kind of "homemade port" of Bioshock from UE3 to Cryengine3? ... Something like that..

If, someday, I have a game development studio / porting studio, I will use this technique.. Maybe it will take long, but the final performance will be better.

As far as I know, there are two kind of porting contracts:

1:The original publisher of whatever game hires an studio for to make a Linux port...

2:An studio / publisher wannabe acquires the publishing rights for Linux of whatever game from the original publisher.

I don't know why, but linux porters like Feral and Aspyr refuse to clarify the situation.
Do they have the publishing rights for Linux of the games they ported, like the second option, or they are just contracted companies, like the first option?

Driver issues are very common.

Whether it be resolution, frame rates...., different hardware/operating systems change the result. For example the ps4 and Xbone's are capable of more limited resolutions and frame rates but the Linux port can actually achieve greater performance that is very much noticeable to the player(4K res, 100+ fps whereas many console games are fps locked low). So from that perspective some of Feral/Aspyr/VP's ports would exceed those standards, just not compared to Windows. On the same note if you port from Windows/Linux to console you will almost certainly lose something. Windows has the advantage of better hardware support (drivers), a larger customer pool and a larger number of devs with experience on the platform. In that regard it is logical the games perform better, and it will take time for that change. It won't change without organizations like Feral/Aspyr/VP. They need to fill the gap so the actual game makers (and more 3rd party devs) can acquire the abilities to do in house Linux games. Even inhouse devs have trouble (Dead Island and Dying Light, the Metro games).

I think the contracts can vary substantially from port to port. But obviously the porters are limited to what the the company giving them contract allow. And the port company's can't force content owners to allow ports they don't want. They can't buy publishing rights when they aren't for sale, or the cost is too high.

We basically agree on most points, except for the pretty major point of whether what the porting companies are doing is good or bad for Linux gaming.

I used Wine in the past and I've dual booted in the past. I've had some games work better in Wine than in Windows. But it is a real PITA, you can't just buy a game, install it and have it work. There are lot's of hoops to jump through, different settings for different games(even different versions of wine to maintain). I'll take a Feral/Aspyr/VP port over Windows or Wine any (and every) day.

First of all, this discussion is completely off-topic, but I will jump in for my 2 cents.

@edddeduck_feral I want to say thank you. In general. To you personally for being here, to your company for what it did so far. I agree with @m2mg2:

Quoting: m2mg2I don't think Feral is bad for Linux gaming at all, I think they are one of the best things we have going for us.

Regarding other discussion, I don't want to focus on comparison of Tomb Raider versions that are hard to compare. I want to focus on one sentence.

Quoting: Comandante ÑoñardoFor me, porting a game to Linux is to take a game that was made previously for another plattform, such as Windows or P$4, and remake it using another game engine

@Comandante Ñoñardo I guess term "game engine" is either something you don't understand or you don't know what world you live in.
Sorry for such a brutal words, I don't want to offend you. I just want to say that from the business perspective it makes no sense to rewrite a game to other game engine while porting it (Except "Enhanced versions" or "Remasters", that focus on improvements, but in this case it's very often update to newer version of previously used engine or even just new compilation of the same in-house engine that was updated already by the team).
Porting focuses on rewriting a layers of the application that are incompatible with target platform. Performance improvements are done to achieve a performance that is satisfying. It's naive to believe that a porter will rewrite a game to other engine while porting a game. Imagine how many issues on how many levels it can bring.

What's also very important - sometimes porting is done by 3rd party company, like Feral. I don't know how much do you know about programming, but just try to fill their shoes - guys receive a code, that was written by other company, that may have different approach to code than all the games they ported so far and they're doing their best to bring the same game to new environment. You would like to add yet another layer of complexity and use new game engine? With different API, with different functionalities, with different tools. Do you know how harder would it be?
I can imagine that XCOM2 was the best experience for Feral, because they could cooperate with game developer and they had influence on some parts of the game code in earlier stages, however I don't know if they started to cooperate from beginning or if they jumped in before the final release(And they had to hurry to be able to release the game at day one).

Anyway, to sum up, I am grateful for everything Feral did. Also it's worth to mention that I admire work done by VP and Aspyr.
EOT.

Quoting: Comandante ÑoñardoI think We have a different concept about game porting.
For me, porting a game to Linux is to take a game that was made previously for another plattform, such as Windows or P$4, and remake it using another game engine 100% native of the target OS (in this case, a native OpenGL linux engine), but with the same graphic assets, same sounds, same game design, same gameplay, same maps, same cinematics, etc.. The final user must not notice any audiovisual difference between the original and the Linux remake...

Converting from one engine to another one is a huge and impractical job. Not even original development teams working on multiple platforms would do this as you're at least doubling your work and will end up with games that behave a little differently on both platforms. Instead we convert the game engine to work on Linux (and Mac) natively so we can best match the original experience to the last pixel, if you start to get into deep semantics then the game engine was designed for DirectX so there are sometimes core design choices that mean the game engine will be more optimally laid out for DX compared to GL these cannot be avoided unless you rewrote the entire engine from scratch which is not feasible however this doesn't mean the game isn't native it just means the design choices are more biased towards the more popular platforms which can sometimes make an impact to the title in certain situations.

For example in Tomb Raider the performance is usually in a similar range to Windows however in certain areas the performance will drop in comparison. In this case it's because draw calls in OpenGL are slower than making the same draw call in DirectX. In the areas like Mountain Village and Shanty town you'll suddenly increase the draw calls from a few thousand to around ten thousand. On Windows this is a negligible hit but on OpenGL the overhead of the draw call will add up and make more of an impact. This can impact some drivers even more than others if they are not as efficient at that call.

We're can be talking about differences of 0.1ms per draw call but something like this could impact the overall frame rate by a massive amount depending on it's usage in the frame, as you can be using a single feature thousands of times per frame.

Now we can do things to optimise the situation by caching states, modifying how certain draw calls are bundled up and many other small tips and tricks. Often these can take months of engineering to get what some people might think is only minor changes. Part of this is working with driver/OS teams (both open and closed groups) to find improvements to gain performance. A large number of the performance improvements in Mesa recently have come from our newer games (and other companies games) exposing performance issues in the drivers.

For example we exposed a performance flaw in the memory manager for Mesa drivers that meant you get pauses and slowdowns (I think Liam posted about this one). Now the improvements to this single area once released could almost double the performance on some AMD cards in certain areas of games! Windows is a much bigger platform so has had a lot more money and engineering time spent on optimising the drivers to get every last bit for performance for games, on Linux the platform is smaller and there have not been many games that really push the performance of the latest features so there is still more potential performance left to find.

That's not saying all the performance issues are drivers no more than all the performance issues are in the game code. My point is bringing games to a new platform especially one that has not had AAA games pushing the graphics to the limits before is not a simple problem and there are so many factors that can impact performance and stability. This is why some games like XCOM 2 or Empire Total War can in some instances out perform Windows whereas others might have some lower performance areas that are below Windows. It's all about the specific problems encountered, features the game uses and the availability of alternate features or solutions to create an optimal solution.

We end up spending more time on the games that have performance deltas than the ones that don't, so if you play a game from us and an area slows down it's likely we spent more time investigating that issue than a different game that runs great on your machine. It's just people assume the opposite as it doesn't perform as well. However if you play our games using the recommended settings on supported hardware/software then you will always get a good experience even if you might get a higher frame rate on Windows. If the experience is impacted then that's something you should contact support about.

We do read forums but if you only mention issues in forums its possible we'll miss your issue as we just didn't read the thread you posted on! So please remember to contact support as that means we'll definitely be aware of your issue.

Quoting: Comandante ÑoñardoI don't know why, but linux porters like Feral and Aspyr refuse to clarify the situation.
Do they have the publishing rights for Linux of the games they ported, like the second option, or they are just contracted companies, like the first option?

All of our game pages has Feral listed as the publisher and the developer on Linux (and Mac). I would have thought that was pretty clear? :)

Quoting: wojtek88I don't know how much do you know about programming, but just try to fill their shoes - guys receive a code, that was written by other company, that may have different approach to code than all the games they ported so far and they're doing their best to bring the same game to new environment.

Yep, as I mentioned above it can be a challenge but it's one we all enjoy taking on every day :)

Quoting: wojtek88First of all, this discussion is completely off-topic, but I will jump in for my 2 cents.

@edddeduck_feral I want to say thank you. In general. To you personally for being here, to your company for what it did so far. I agree with @m2mg2:

Quoting: m2mg2I don't think Feral is bad for Linux gaming at all, I think they are one of the best things we have going for us.

Regarding other discussion, I don't want to focus on comparison of Tomb Raider versions that are hard to compare. I want to focus on one sentence.

Quoting: Comandante ÑoñardoFor me, porting a game to Linux is to take a game that was made previously for another plattform, such as Windows or P$4, and remake it using another game engine

@Comandante Ñoñardo I guess term "game engine" is either something you don't understand or you don't know what world you live in.
Sorry for such a brutal words, I don't want to offend you. I just want to say that from the business perspective it makes no sense to rewrite a game to other game engine while porting it (Except "Enhanced versions" or "Remasters", that focus on improvements, but in this case it's very often update to newer version of previously used engine or even just new compilation of the same in-house engine that was updated already by the team).
Porting focuses on rewriting a layers of the application that are incompatible with target platform. Performance improvements are done to achieve a performance that is satisfying. It's naive to believe that a porter will rewrite a game to other engine while porting a game. Imagine how many issues on how many levels it can bring.

What's also very important - sometimes porting is done by 3rd party company, like Feral. I don't know how much do you know about programming, but just try to fill their shoes - guys receive a code, that was written by other company, that may have different approach to code than all the games they ported so far and they're doing their best to bring the same game to new environment. You would like to add yet another layer of complexity and use new game engine? With different API, with different functionalities, with different tools. Do you know how harder would it be?
I can imagine that XCOM2 was the best experience for Feral, because they could cooperate with game developer and they had influence on some parts of the game code in earlier stages, however I don't know if they started to cooperate from beginning or if they jumped in before the final release(And they had to hurry to be able to release the game at day one).

Anyway, to sum up, I am grateful for everything Feral did. Also it's worth to mention that I admire work done by VP and Aspyr.
EOT.

To echo this post thanks for being here.

Also it's clear some people don't quite understand what porting is yet.

You don't move a game to a different game engine, that's insanity for a porter and could end up taking a very long time.

Excellent and informative post edddeduck_feral, you should do this more often when people here or elsewhere start lashing out at Feral.
I also hope that you don't get discouraged by these types of commentaries and you'll continue to do your outstanding work.

Quoting: edddeduck_feralConverting from one engine to another one is a huge and impractical job. Not even original development teams working on multiple platforms would do this as you're at least doubling your work and will end up with games that behave a little differently on both platforms. Instead we convert the game engine to work on Linux (and Mac) natively so we can best match the original experience to the last pixel,

That's the kind of answer I want: Technical and specific. ^_^
What are the technical impediments for to bring Vulkan support for this game (at least, as an experiment)?

Quoting: edddeduck_feralFor example in Tomb Raider the performance is usually in a similar range to Windows however in certain areas the performance will drop in comparison. In this case it's because draw calls in OpenGL are slower than making the same draw call in DirectX. In the areas like Mountain Village and Shanty town you'll suddenly increase the draw calls from a few thousand to around ten thousand. On Windows this is a negligible hit but on OpenGL the overhead of the draw call will add up and make more of an impact. This can impact some drivers even more than others if they are not as efficient at that call.

We're can be talking about differences of 0.1ms per draw call but something like this could impact the overall frame rate by a massive amount depending on it's usage in the frame, as you can be using a single feature thousands of times per frame.

Now we can do things to optimise the situation by caching states, modifying how certain draw calls are bundled up and many other small tips and tricks. Often these can take months of engineering to get what some people might think is only minor changes. Part of this is working with driver/OS teams (both open and closed groups) to find improvements to gain performance. A large number of the performance improvements in Mesa recently have come from our newer games (and other companies games) exposing performance issues in the drivers.

That's is what I want: Optimizations for those areas...
I don't care if I have to wait another year; I prefer a delayed port with a sustained 60 fps rate using the same hardware of the windows version, than a day 1 port with huge performance drops... (Remember the Linux DEMO of the new System shock?)

Quoting: edddeduck_feralThat's not saying all the performance issues are drivers no more than all the performance issues are in the game code. My point is bringing games to a new platform especially one that has not had AAA games pushing the graphics to the limits before is not a simple problem and there are so many factors that can impact performance and stability. This is why some games like XCOM 2 or Empire Total War can in some instances out perform Windows whereas others might have some lower performance areas that are below Windows. It's all about the specific problems encountered, features the game uses and the availability of alternate features or solutions to create an optimal solution.

For that, We need a (timed)Linux exclusive AAA game... But there aren't developer teams with the balls and/or skills for that... The money problem can be solved through Crowdfunding (but if you want money, you must have a playable DEMO.)

We only need just one AAA timed Linux exclusive for to make the difference.

Quoting: edddeduck_feralAll of our game pages has Feral listed as the publisher and the developer on Linux (and Mac). I would have thought that was pretty clear? :)

So... If feral is the publisher, I guess that Feral has the legal power to publish the Linux and mac version of this game on other stores, like GOG.. Maybe not this year, but in 2017.

Now, two Off topic technical questions:

Instead of use Direct3D-»OpenGL API translation layers (thing that eats a lot of CPU resourses, like Wine), why not implement Direct3d directly on Linux? What are the technical impediments for that?

Why Nvidia Physx by GPU for Linux wasn't implemented yet?

I've said this elsewhere, but I'm guessing Feral sees if they can get a relatively sustained 60fps across a range of settings and hardware and if they can it's good enough for release. Note that when you played the Tomb Raider port initially on a lot of hardware you'd be able to pull over 90-100fps on High @1920x1080. At that point there's not a reason for them to delay release as long as its not buggy. Then they went ahead and had people like me and I'm sure others in the community contact them when the performance delta was off in parts of the game and send them lots of data which they were able to use to pinpoint ways to improve performance even further.

And I guarantee that without the extra feedback from its release the game would have languished for a long time unnecessarily. At the end of the day we'd like to be able to play the games, and Feral would like to release them and get paid.

Also I'm sure the basic API calls aren't the overhead issue so much as the difference in how GL and D3D handle draw calls. In GL everything has to be batched because you only get a single context to draw things in. Additionally the way D3D games expect to load shaders causes a stall in GL because for GL it's all compiled at run-time (the first time it's used) and for D3D it's all pre-compiled. That's what edddeduck_feral was talking about. There's not much way around it past a certain point without gutting large enough portions of the engine to not be cost effective. I'm sure in the case of XCOM2 (for example) they were there early enough in the process to work with the original developer and keep it from being as much of an issue. But that's not always going to be the case.

I'm sure something that doesn't help as well is the translation of HLSL shaders to GLSL. Assuming you aren't (maybe you are) rewriting every shader from scratch you're going to miss out on a ton of optimizations. But again the time involved would make it not cost effective.

You have to remember it's a studio of something like 70 people currently working on something like 6 or 7 major titles (and usually for two platforms), in addition to providing support and patches for titles already released. They can't just throw people and time at every game.

Quoting: Comandante ÑoñardoThat's is what I want: Optimizations for those areas...
I don't care if I have to wait another year; I prefer a delayed port with a sustained 60 fps rate using the same hardware of the windows version, than a day 1 port with huge performance drops... (Remember the Linux DEMO of the new System shock?)

And who is going to pay wages for the whole team during that year? Will the port be still profitable if you attach another $$$$ bill to it?

We're not talking about amateur enthusiasts here but employees with families and mortgages.

Quoting: Comandante ÑoñardoInstead of use Direct3D-»OpenGL API translation layers (thing that eats a lot of CPU resourses, like Wine), why not implement Direct3d directly on Linux? What are the technical impediments for that?

There's no common back-end to write the interface on. There is Gallium Nine, which has DX9 implemented onto Gallium. But this only works for (some?) open source drivers. For proprietary this can only be done by the driver developers. Which means that you can now choose to make a faster implementation on slower drivers or a slower implementation on faster drivers.

@edddeduck_feral, thanks for the interesting read. That answers some of a very long list of questions. :P

Why run the game through Wine/Crossover when the native game runs just fine? I'm using a GTX970 with Nvidia 354 Driver on Ubuntu 16.04. 20% into the game and no visual issues so far, only minor stutterings in some areas, but very rarely. Overall performance, especially in combat is just on point, it runs smoothly. So where is your point running this game via Wine/Crossover for more performance, when performance is already there in native version?

Offtopic: Why no D3D on Linux? Because lack of source availability of D3D on M$'s side, and M$ is not interested in porting D3D to their biggest competitor. There is also no need for it. OpenGL (and Vulkan) is there for everyone to use royalty free. Game Developers just need to use OpenGL instead of D3D, even for Windows.

Quoting: Comandante ÑoñardoWhy Nvidia Physx by GPU for Linux wasn't implemented yet?

As far as I know this game doesn't use hardware accelerated PhysX on any platform. It just uses standard software mode.

@edddeduck_feral

Great description! I really don't understand why people even compare the game performance over different platforms so much. It is nice to have a reference, but if the game runs at 250fps on dx9 and half of that on openGL I don't really mind that because it is still playable and every game that I'd tried have worked pretty well (I do have gtx970 though). It is not a secret that game that is written for dx9 or dx11 will not run the same on openGL but we cannot expect it to and we should be happy for every game that we do receive for a marketshare so small, and not complain about having worse performance than some other OS, however it might be frustrating for people with lower-end hardware. We all have choice of choosing our daily drivers after all, I personally would rather pay more for hardware part and actually own a computer and OS rather than having Windows 10 installed :) It is really great what feral and similar companies are doing for our community and I cannot thank enogh for providing enough games so that I could ditch Windows once and for all :D Hopefully we will be getting just as many titles and even more constantly and hopefully we will see Vulkan API used in future games!

Quoting: edddeduck_feral

Quoting: Comandante ÑoñardoWhy Nvidia Physx by GPU for Linux wasn't implemented yet?

As far as I know this game doesn't use hardware accelerated PhysX on any platform. It just uses standard software mode.

That was an offtopic question about the current status of physx on Linux...

I was about to make the experiment of Tomb Raider with wine-staging and gallium9, just for to see what happen... but it seems that only Radeons are supported by Gallium9... and Manero already did the job with[Tomb Raider](https://manerosss.wordpress.com/2016/08/03/tomb-raider/) and [Life is Strange](https://manerosss.wordpress.com/2016/07/25/life-is-strange/)

About Gallium, the idea of to teach Linux how to speak Direct3D instead of using a D3D--» GL layer is revolutionary.

Anyway, all this was about scientific curiousity....

My Win7 gaming machine is almost ready... My gaming life is this close the get more rich... You know, the windows games catalog is SO big... Instead of waiting years, soon I will be able to play games on day 1 release (or in the first 50% off sale)... ^_^

Quoting: MaCroX95t is not a secret that game that is written for dx9 or dx11 will not run the same on openGL but we cannot expect it to and we should be happy for every game that we do receive for a marketshare so small, and not complain about having worse performance than some other OS, however it might be frustrating for people with lower-end hardware.

My point exactly...
In countries like Argentina, the hardware costs the double than in USA... Most people barely can afford a low-end or mid-range GPU's and CPU's... and that people will want the best FPS performance per the few dolars they can spend... and that just can not be achieved on Linux yet.

GPU Nvidia Physx is already available for Linux since 2014. I think Metro Redux games are using it.