Former Valve engineer Rich Geldreich has written up a blog post about the state of Linux Gaming. It's an interesting read that's for sure.
When talking about recent bigger game ports:
QuoteSadly, it's pretty clear that if you run these games on Linux your experience isn't going to be as good, and you'll be getting less "gaming value" vs. Windows. We're not talking about a bunch of little indy titles, these are big releases: Borderlands: The Pre-Sequel, Borderlands 2, Tropico 5, XCOM: Enemy Unknown, Sid Meier's Civilization V. My take is the devs doing these ports just aren't doing their best to optimize these releases for Linux and/or OpenGL.
Emphasis mine, and I don't agree with him on this. Obviously neither he, nor I have any proof either way that they are/aren't doing their best to optimise, but Aspyr & Feral making a living out of porting games to OpenGL, so why wouldn't they be trying to fix performance issues?
The performance has some way to go sure, but is that really the fault of Aspyr & Feral, or do the drivers still have ways to go to improve their performance? Who knows, I sure don't it goes way over my head at that point.
He does however note how hard it is to get performance on Linux equal to Windows:
QuoteI know it's possible for Linux ports to equal or outperform their Windows counterparts, but it's hard. At Valve we had all the driver devs at our beck and call and it was still very difficult to get the Source engine's perf. and stability to where it needed to be relative to Windows. (And this was with a ~8 year old engine - it must be even harder with more modern engines.) These devs are probably glad to just release anything at all given how alien it can be for Windows/Xbox devs to develop, debug, and ship stuff under Linux+OpenGL.
At least he is pointing out that fact that it's not easy to get decent OpenGL performance to match games on Windows, so he's not completely blasting Feral and Aspyr.
I agree with what he's saying about the Intel drivers 100%:
QuoteThe entire Intel driver situation remains in a ridiculous state. I know Intel means well and all but really, they can do better. (Are they afraid of pissing off MS? Or is this just big corp dysfunctionalism?) Valve is still paying LunarG to find and fix silly perf bugs in Intel's slow open source driver. Surely this can't be a sustainable way of developing a working driver?
No, it's not. Intel ideally needs to be doing this sort of work themselves to find bottlenecks and fix general performance issues in their own Linux drivers. I see this as a stopgap measure while Linux still isn't a focus for many people, and Intel included. This again goes into our marketshare issues, if we had a higher share then Intel would probably be doing it themselves.
His last point is a screen-shot of a slashdot comment where someone is basically saying that SteamOS is done, and that we will never get our hands on the Steam Controller. Their reasoning is that Microsoft snapped and allowed Alienware to create a Windows machine that boots to a Steam UI, and not Windows directly.
I agree that was a bit of a shocker, and I thought it wouldn't do SteamOS any good, but I think SteamOS hasn't even had a chance to have a go at it yet. SteamOS was delayed because Valve decided to revamp the controller again to get it right, so I think we should wait for it to be out before signalling its death.
Read his full blog post here.
What are your thoughts on it? Rich is good at stirring up the hornet's nest that's for sure, but just because he is a former Valve engineer doesn't mean he's going to be right on everything. Article taken from GamingOnLinux.com.
0 Likes
Some you may have missed, popular articles from the last month:
The comments on this article are closed.
Quoting: AstroQuoting: FutureSutureYou must be a troll.I'm not a troll.
Then stop trolling!!
2 Likes, Who?
Quoting: ImantsStop accuse me of trolling. It's obvious I'm not doing that.Quoting: AstroThen stop trolling!!Quoting: FutureSutureYou must be a troll.I'm not a troll.
0 Likes
I personally think there are quite a lot of valid things in there. I don't think that Aspyr/Feral are "not doing their best" on the games, but that it simply is hard, and optimizing is hard. Fact is, that our graphics stack - especially x11 - is aged, even though, sdl2 handles a lot of the hard x11 input stuff pretty well.
A fact is, that we do have problems with our stack, in drm, in drivers, in mesa, in X11, supporting more than nvidia is about working around breakage in the drivers, and we know amd ones and intel ones are pretty borked (for years now).
Another hard thing is - look on the windows side. Even there, OpenGL is broken almost in every driver. If OpenGL was a real option for game development in Windows, we probably would see better implementation in the beginning and maybe game developers would even consider switching to OpenGL.
You reach a market share of 1 % by Linux, so you may not invest the same time tuning the thing as on windows, where you currently reach a market share of about 95 % of the (PC) gamers.
I personally don't think we'll have that issue a lot in the future. We'll see OGL4, and we'll see DX12. And within that, we'll see more Azdo implementations with DX12/OGL4 accessing lower level features rather than using high level methods in future. Game engine developers wanted that for a long time, and now we're there. Don't get me wrong, OGL/DX are good enough for a lot of things, but game engine developers who want and need to push the best out of the hardware - I guess that's where we're going now.
A fact is, that we do have problems with our stack, in drm, in drivers, in mesa, in X11, supporting more than nvidia is about working around breakage in the drivers, and we know amd ones and intel ones are pretty borked (for years now).
Another hard thing is - look on the windows side. Even there, OpenGL is broken almost in every driver. If OpenGL was a real option for game development in Windows, we probably would see better implementation in the beginning and maybe game developers would even consider switching to OpenGL.
You reach a market share of 1 % by Linux, so you may not invest the same time tuning the thing as on windows, where you currently reach a market share of about 95 % of the (PC) gamers.
I personally don't think we'll have that issue a lot in the future. We'll see OGL4, and we'll see DX12. And within that, we'll see more Azdo implementations with DX12/OGL4 accessing lower level features rather than using high level methods in future. Game engine developers wanted that for a long time, and now we're there. Don't get me wrong, OGL/DX are good enough for a lot of things, but game engine developers who want and need to push the best out of the hardware - I guess that's where we're going now.
0 Likes
FutureSuture Nov 11, 2014
Quoting: AstroYou are. You very clearly are. Read the bit in bold again. How is sabotaging the efforts of those few developers and publishers who do support us not trolling? Tell me. How is ruining this endeavour, one that was unbelievable to ever happen a mere few years ago, for the vast majority of us not trolling? Tell me.Quoting: FutureSutureI'm not a troll.Quoting: AstroGamingOnLinux has had a spotless community in my eyes so far. That changed for me when I read this comment, especially the bit in bold. Please do not post here again unless you stop being so completely and utterly obnoxious. I mean, seriously, what the actual fudge? You must be a troll.Quoting: BeamboomThis right here is what I talk about in my post above. Ferals work on Xcom is the exact opposite of garbage, it's one of the best ports to Linux so far, and to sit an spew out stuff like this on forums on the net... We're just in no position whatsoever to do so!A lot of people contact me after reading my comments about Feral. We talk and I recommend them contact Feral and ask them don't port any more games on linux. So, I don't need to argue with you here. Sorry, nothing personal. What I need is just leaving my comments and do what I do.
0 Likes
Trolling is when you do or post something outrageous to elicit a reaction. I think he's just an angry customer. He doesn't owe anything to anyone. If he wants to complain about the port not supporting 32 bits when the Windows version does, it's his right. If he wants to do what angry customers do and tell his friends, it's also his right. You may judge him for the reason he's angry, but he still can do all these things as a customer. Customer rights trump the goal of making linux a gaming platform, as far as I'm concerned.
1 Likes, Who?
What does Valve says about the linux game ports?
0 Likes
Mohandevir Nov 13, 2014
The problem is that game development on Linux is a work in progress and Valve has given it a new momentum. Still, there's a lot of work to do and some of us hope to have performances on par with Windows.
We should keep in mind that DX has a 15 years lead of development on us.
I'm very confident that things are going to get better over time but in the meanwhile I understand why some of us are disapointed.
Maybe we should see the state of gaming on linux like a big beta and we are contributing to make it better?
We should keep in mind that DX has a 15 years lead of development on us.
I'm very confident that things are going to get better over time but in the meanwhile I understand why some of us are disapointed.
Maybe we should see the state of gaming on linux like a big beta and we are contributing to make it better?
1 Likes, Who?
>> These devs are probably glad to just release anything at all given how alien it can be for Windows/Xbox devs to develop, debug, and ship stuff under Linux+OpenGL.
There is a lot of test results that show truly equal code generally performs faster under Linux than Windows.
Linux and Windows programming are mostly whole different mindsets.
This is just an educated guess because I've seen it first-hand before, but I'd be very surprised if just the decision to use Windows developers to also write the Linux version didn't alone account for at least most of the negative performance differences, since their first instincts will be to keep reverting to a more familiar (windows) paradigm even if it requires a cpu-expensive abstraction layer to support it under Linux, than just adapt to and adopt the native Linux one.
There is a lot of test results that show truly equal code generally performs faster under Linux than Windows.
Linux and Windows programming are mostly whole different mindsets.
This is just an educated guess because I've seen it first-hand before, but I'd be very surprised if just the decision to use Windows developers to also write the Linux version didn't alone account for at least most of the negative performance differences, since their first instincts will be to keep reverting to a more familiar (windows) paradigm even if it requires a cpu-expensive abstraction layer to support it under Linux, than just adapt to and adopt the native Linux one.
0 Likes
Patreon
PayPal
See more from me