Higher-end AMD cards with the binary drivers don't have any issues that I can tell with any of the Feral Linux ports. If you use the OSS drivers, performance is worse, and recent GL features are not available which can prevent a game from running or rendering properly ( e.g. on a R9-290, Shadow of Mordor renders the scene and character face/hands, but without the rest of the character; amusing but not quite the experience I'm after ).

I'd rather not blame Feral, AMD or Intel without knowing why they don't officially support certain hardware. It's likely that porting each game is different, depending on the original technology used ( engine, middle-ware etc ), and the renderer is just one part of the porting puzzle.

Under Windows, D3D provides a single development target; when there is no GPU support for a particular function, D3D provides a CPU software implementation. Under OSX, Apple insist on providing the only OpenGL stack so that they can implement the same sort of software fallback mechanism. Under Linux, there is no single OpenGL provider, and each GL stack has different feature coverage, strengths and weaknesses.

Perhaps asking Feral for comment on hardware support would be more helpful, but I suspect it comes down to the fact that NVidia has 2/3 of the GPU market, and Valve have worked primarily with NVidia to kick-start the Linux games market. Linux games are still marginal in terms of revenue, so porting costs have to be considered carefully, particularly in cases like this where fragmentation pushes them higher.

View this comment - View article - View full comments

Quoting: mr-eggEven though im an nvidia user it would be nice to get some AMD support here. AMD are going to be improving hopefully soon and its not good if games don't even run when the driver gets a huge uplift.

Hmm, quite a few recent Linux games seem to come with the caveat "AMD not supported", which is a concern personally as I currently use AMD, but also more generally, we want to have decent GPU choice to prevent monopoly pricing and technology stagnation.

Fortunately, I have found that the R9 290 I use seems to work fine with most new games, but there are still one or two that exhibit GPU faults.

View this comment - View article - View full comments

@TheBoss

You shouldn't be concerned about the extension mechanisms of OpenGL and Vulkan. Extensions have always been a major strength of OpenGL, allowing new new features to be used as soon as they are available. Features that developers want ( i.e. actually use ) then tend to find their way into the main specification.

In contrast, with D3D new hardware features only appear when Microsoft choose to expose them with a new D3D release. This meant that some Windows gamers paid for hardware features that were never used, because Microsoft would use D3D features as a business tool to play hardware manufacturers against each other.

Any good developer would want decisions about what hardware to use and how to use it left to them, not decided in the business interests of the OS manufacturer.

View this comment - View article - View full comments

I think NVidia are following the same logic that Microsoft applied with D3D12. Microsoft allow you to write a hybrid D3D11/D3D12 application, where you take advantage of D3D12 API to construct commands on multiple threads/cores of the CPU to get over the draw-call limit that both D3D and OpenGL had been suffering from. The rest of the application continues to be D3D11.

For a game whose performance suffers from draw-call limits, this is a lot less risk and a lot less work than redesigning your renderer, and possibly other parts of your engine to conform to a new development model.

But single-threaded draw-call preparation is far from the only reason why Linux games run slowly. The biggest problem is that a capable PCIe bus GPU is going to be stalled by almost ANY interaction with the CPU because the bus and CPU will both introduce latency. Submitting a draw call therefore has a cost, changing state or state blocks has a cost, and having to send or receive data to synchronise CPU and GPU memory has a much greater cost. These are costs that you can't really avoid if your engine splits processing between CPU and GPU as has normally been the case in the past.

The AZDO API calls added to OpenGL 4.x ( particularly multi-draw indirect and bindless textures/buffers ) along with compute shaders and OpenCL kernels, would in theory allow you to rewrite an engine such that after initial setup, almost all of the work is internal to the GPU. Trouble is, to make that change is a lot more work than you might think. There is only a small body of reference papers describing recent experiments in avoiding CPU interaction, so such a conversion is risky and slow to implement, which is why we have not really seen much benefit from AZDO yet.

I'm guessing that we will have the same issue with Vulkan. It will improve command submission and allow almost complete control over the GPU, but at an even greater development risk and cost than AZDO. I have to say that, apart from increasing draw-call submission, I think it will also be quite a while before game developers really work out the best way to use the benefits of Vulkan.

View this comment - View article - View full comments

Quoting: Mountain ManI've been quite happy with Empire. It's the first Total War game I've ever played, but everything I've read suggests that it's where the series peaked (either Empire or Shogun II, depending on who you ask).

All the Total War games are good in their own way, and they are all at least a little different from their predecessors.

The first 2 ( Shogun, and Medieval ) had some game mechanics that I quite liked. The campaign at world level was a bit like moving on a Risk board, the tactical combat was quite fast and furious at times.

Then there was a significant change, with the campaign map becoming a 3D world with army/navy counters moving distances according to troop type and terrain, with every region able to be enhanced to a huge city. This was a really positive advance at the time, so not many lamented the change. This basic blueprint persisted from Rome through Medieval 2. Empire and Shogun 2 ( plus their expansions, of course ), so it became quite deep-rooted as what a Total War game should be, for a lot of players.

The last game, Rome 2 and its spin-off Attila have adopted two major changes that not everyone like. First, regions are now grouped into provinces, where only one region can grow to a major city, and not all items can be built in all regions. Second, you are limited to the number of army/navy stacks you can have. So a lot of players regard the game as going downhill after Shogun 2. I don't really agree with that; I don't necessarily like all the Rome 2 changes, but overall it hangs together as a game - just one that you need to think differently about.

I wouldn't necessarily recommend the first 3 games, except for completeness, as they are a bit dated now, but if you liked Empire, I don't see why you wouldn't like Medieval 2. I think you just need to treat each game on it's merits rather than expecting cloned features game-to-game.

I am a bit annoyed that the Rome 2 Linux port has disappeared off the radar, as I bought it when they announced the port as a gesture of support ( I won't make that mistake again ), but I was happy that Attila came out recently, and now we have both Medieval 2 and Empire from Feral.

The only Total War game I don't own is Shogun 2, so it's a shame Feral don't have a Linux version of that; I would like to be able to reward their efforts, but already have the games they have ported from when I used to use Windows.

View this comment - View article - View full comments

I also use AMD, not because I have any particular problem with NVIDIA, but because AMD generally provide better hardware for a given price point. NVIDIA have a better overall software package,; they employ lots of engineers to ensure drivers recognise games individually and ensure the driver response is optimal for that game.

Vulkan does away with the complexity of OpenGL, so NVIDIA will lose this software advantage. This will either result in pressure on NVIDIA to reduce their prices, or will allow AMD to raise theirs. I know which I would prefer.

My current card is a stock R9 290, so decent, but not tthe latest. It runs every new Linux game I try at high settings, including Shadow of Mordor and Total War Attila which don't supposedly support AMD. I haven't tried my old 5850, because I am sure it would fail miserably on new games. But I am sure it would fare no worse than NVIDIA cards from that generation. This is just how it goes, you can't expect your hardware to keep up beyond about 3 to 4 years, because new games always wan't and expect more power.

View this comment - View article - View full comments

This is not particularly a surprise. Khronos actually has a decent record of getting things right since they took over OpenGL. I think they are aware that they need to have existing implementations available at release with no inconsistencies between them. Whether it is available now or some time q1 2016 does not really matter until game back-ends are recoded to use Vulkan.

I am more interested to see what Apple, Sony and Microsoft try to do. It is important long term that Android support Vulkan as it is the largest open platform, but the vendors more inclined to lock their platforms could cause problems by preventing Vulkan implementations or providing only bad implementations on their platforms.

View this comment - View article - View full comments

This is probably NOT completely an accident. If you look at M2TW at https://steamdb.info [External Link] you will see that it has configurations and depots for Windows Mac and Linux, and has done for a few weeks.

I expected it to be released at some point, but was not sure if the port was in-house or 3rd party. Looks like it is 3rd party like ETW. They are probably just not finished with the Linux version yet, but I think we will get it eventually

Its possible that CA have a long-term plan to port everything to SteamPlay, but don't necessarily want to promise anything, based on their experience with R2TW :)

View this comment - View article - View full comments