Slashdot Mirror
Expert Opinions On Linux Gaming's Future
jg21 writes "Following
on from yesterday's Slashdot coverage of the idea to launch a games-based Linux
distro, LinuxWorld Magazine has held a Gaming Round Table involving Chris DiBona, Ryan
Gordon, Timothee Besset, Gavriel
State, and Joe Valenzuela about where Linux
currently stands and how it will one day become a premier gaming platform. 'It
became perfectly clear to me that most of the technological issues are already
solved, and that the others won't take too long to fix once the game publishers
really get into the mix,' reports Dee-Ann LeBlanc, Gaming Industry Editor for
LinuxWorld, who coordinated the round table. Well worth reading."
It's nice to talk about creating a "gaming OS", but the key component here is that you need some games.
Sokoban and Mahjongg only get you so far..
OpenGL exists on Linux, what else are game developers missing?
I have been pwned because my
I wonder if they contacted John Carmak about this... or even concidered him. I mean, he and his team did create the first true 3d (raltime) game (wolfenstien, for those you living inside a cave) and his company does support Linux (Quake III Arena, for example)
The most annoying thing is getting the grafix drivers to work properly. When I was trying to get UT2003 to work, I found the install to be the easy part, but finding the proper drivers and installing them was the most difficult part.
If the dollar is an "I owe you nothing", then the Euro is a "Who owes you nothing." - Doug Casey
How can I play a game when I'm blind?
YOU INSENSITIVE CLOD!
Nethack has pretty good support for such technologies as screen readers and braille pads.
"They redundantly repeated themselves over and over again incessantly without end ad infinitum" -- ibid.
Traditionaly the gaming industry is one where garage developers have great impact.
A big problem I see with Linux as a mainstream gaming platform is that there is no significant market to tempt those developers with no extra money to burn...
I speak myself as a former game developer (now on the academic side of the world)... how would you convince me to develop for linux if I have no extra money??
--krahd
mod me up scottie!
One thing that Linux can do really well is CLUSTER EASILY. Forget the PS3... as long as games are written to make use of Linux's clustering abilities - we can have some MASSIVE gaming servers and game environments.
:)
Now all I want for Christmas is an Open Mosix release for the 2.6 kernel.
READY.
PRINT ""+-0
Obviously, we've all heard about UT2003/2004, Neverwinter Nights, and the upcoming Doom III (id Software usually supports Linux well, yay them!)
Even the US gov't is jumping aboard with America's Army (as well as support for Mac).
Linux is growing, and needs to grow more and more in regards to users, so we can get better game AND hardware support. I know some people think this Linux vs. Windows war is kind of silly, but until Linux grows to the point where it's recognizable by the average user we'll still be left out in the cold in many regards (such as, of course, games and hardware).
I admit, I myself still have Windows installed. How else can I play many games? Wine doesn't want to work on my computer, and it's not perfect anyway.
---
Never criticize religion on Slashdot. You will be modded down for "Troll" no matter how factual it is.
The most important thing required for a successful gaming platform is an audience to purchase the games. If you have that, game developers will develop for your platform. Linux does not have this yet and it is a bit of a chicken and egg scenario.
About the only thing Linux can hope for in the short term is the occasional port but even that may not be financially viable for quite some time judging by the smouldering crater that was once Loki.
Hmm...I wonder if it'd be easy to convert Mac OS X games to Linux? After all, both Mac and Linux games use OpenGL, and both Mac and Linux are UNIX based... If the developers take the Mac source code and tweak it a bit for linux, then recompile it on an x86 Linux machine, voila, Mac games on Linux!
Two Words: Market share.
The games will come if/when a larger proportion of their target market runs Linux.
Right now, very few games are developed for Linux, because relatively few game buyers run Linux. Most game developers don't have the time or resources to port their products, because the margins are razor thin and time is critically important. Windows development toolkits like DirectX are widespread and proven effective.
Until linux is percieved as a major market and has the level of (hardware) vendor support that Windows-based stuff does, it will continue to be an afterthought in game development.
"We have to go forth and crush every world view that doesn't believe in tolerance and free speech." - David Brin
Personally I think it might just be a bit easier to roll out a gaming linux console, as it eliminates most of the installtion/setup processes that could be complicated sometiems.
Rock that crushes, Paper & Scissors that don't matter.
Linux seems to be stuck right now as far as games go. There are GREAT free games, don't get me wrong, I've wasted many hours using frozen bubble, but, there needs to be incentive (read users) for commercial game developers to develop for linux. The catch-22 is that there needs to be incentive (read games) for windows users to switch to linux. I'm not a big gamer, so doesn't effect me and I'd rather buy a game console, however, joe six pack needs games that can play easily on his OS, before a switch.
The FA makes some valid points about the cost of porting games to linux. However, there are commercial-quality game engines out there that do run under linux. One of them, Nebula if even open source (even though Nebula2 is still lacking graphics support for linux, but that's in the works). Nebula1 is perfectly useable and has all kinds of goodies, including input handling, sound, and a slick architecture.
I believe the major problem at the moment is definitely the difference in availability/quality of hardware accelerated graphics drivers. One ATI get their shit together, the story might be different...
I asked for a refund - and got my monkey back.
State: It's very simple. Buy more games and tell the industry that you're buying that game to play on Linux.
I totally agree. The single biggest hinderance to seeing more games running natively on linux is the perception (and likely fact) that there's no money in it. It's for this reason that I subscribe to Transgaming, Bought Neverwinter Nights (and sent them a letter explaining why I picked their game and thanking them), and have copies of games from (some defunct) companies that I dont even play, but whose development I thought it was important to support.
Just keep supporting the folks doing a good job.
---
Jedimom.com, picking out a thermos for you.
StrategyTalk.com, PC Game Forums
I posted this in the last Linux - Gaming - Distro - Thread, but was a bit late. This isn't consistent in itself, but the idea should be clear:
What about setting up a fund for developing a linux game? It should have a concept, only rough, like the genre, set.
Then set up a website with a nice progress bar, and a target sum needed for the developement, like what? 5 Million Dollars? 10 Million Dollars?
Ok, that won't get us a completely new Half - Life - 2 developed, but maybe a nice RPG / Adventure built on an existing engine.
Maybe different Funds for different uses, like
- Make a cool RPG a 'la Deus Ex / System Shock
(Wizardry would be even better, but i don't know about the mass - marketing appeal...)
- "Make a good game developing environment based on Crystal Space"
Make an agreement with some game studio to get a cool engine for a guaranteed price for a free - as - in beer - game production use, let it be the UT or Doom 3 Engine. Or not, depends on the game's genre, i guess.
Let somebody develop a cool game from this money for the community.
If the community wants a new cool game developed, everybody transfers a few bucks to a new proposed game fund of his choice. I think there are enough gnu / linux / bsd / mac etc. fans out there to invest a few dollars each to get a big enough budget, it's mostly a marketing question, i guess.
Kind of like the effort for opening the Blender source?
The fund should be handled by a trusted entity, of course.
The biggest problem is convincing developers that there's money in it for them.
Most are under the impression that they shouldn't bother with anything other than Windows because there's no money in it. "95% of the market is Windows, so why bother with a poultry 5%" type attitude.
Also, added to the cost is desktop support. If you write a game for just Windows you only have to worry about Windows problems. If you write a game for Linux and Mac OS X, you have to hire, train, and then troubleshoot Linux and Mac problems.
The other problem is to convince developers to NOT design their game around proprietary technologies such as DirectX.
By the way, this information comes from the developers themselves. Personally, I think it's a bunch of crap excuses for lazyass companies trying to squeeze out every profit they can by minimizing responsibility. I'm an avid Mac user but I just recently had to buy a PC just to play games. Counter-Strike, Infantry, and Subspace are Windows only and impossible to play under emulation. However, I'd LOVE to see all my favorite games running under Linux and Mac OS X so I can chuck Windows.
If game developers can't be convinced to even write games for the Macintosh using the above excuses (especially the marketshare one), why would they be at all interested in a desktop that has an even smaller marketshare than the Macintosh?
While we tend to blame the problem on Linux's small marketshare, I think Ryan is right here in that binary compatibility has as much, if not more to do with it. Compared to Windows, it would seem that things get broken more often in Linux, both application and driver wise, and that no one from the glibc guys to Linus himself want to really support this kind of compatibility in fear that it will undermine the OSS movement. How is an industry that needs binary compatibility for games and drivers alike supposed to survive without it?
...Linux is a pain to develop a good game client for. DirectX games are not easily ported, and most games are DirectX. This means most professional game developers are fluent in DirectX. DX makes things a lot easier than writing for every sound/video card out there.
Further, Linux editions of games lose money. Quake3 for Linux sold dismally, while people were buying the Windows version enough to be dunking the CDs in their coffee. And the Linux client was released first: if ever there was an opportunity for a killer-app game to help boost Linux, that was a great time.
Loki went out of business by doing the smart thing: bootstrapping itself with porting Triple-A titles from Windows, to earn some cash and develop a library to live on. Who's going to look at the Linux market and see it as viable when id and Loki can't make a good go of it?
And Linux users are habituated to not paying for Linux software, as a rule. Not that they don't, and not that there aren't vertical markets where people are paying good money for Linux apps, but the OSS community is, well, a hard community to pry money out of.
I say this as a developer of Windows games, who runs Macs at home and who has compiled a few Linux kernels in the past. Developers have enough to do to create a modern game while taking advantage of the assistance of things like DirectX: taking on the burden of developing the same thing without that help, for a community that likes their software free (both kinds of free),... that's a lot to expect.
Honestly, if I were a games developer looking at the Linux market seriously, there is one feature which would really draw me in: the ability to provide a bootable distribution on the game CD.
One of the biggest headaches of game developers is trying to test their game on a sufficiently large subset of available hardware and software configurations to insure it will work properly. This isn't an issue on Consoles, which is one (not the only, but a big one) of the issues they are so popular to develop for. Having a bootable distro on the game CD gives the developer many of the advantages of both Console and PC:
Given the size of modern games, DVD distros are more likely than CD distros, but the concept is identical.
The bootable game CD/DVD has the potential to drastically reduce developer costs associated with modern games, and merge the best features of PC and Console gaming, with few drawbacks. I expect to see game makers venture into Linux in this area first.
-Erik
There are always four sides to every story: your side, their side, the truth, and what really happened.
As many have said on the previous discussion, games-oriented distros already exist, based on Knoppix, Gentoo and many others. I can't help but feel the focus on these distro developers is not going to the right direction.
Being a developer myself, having used UNIX clones for more than one decade, and worked in the videogames industry, I know it's tempting to see the whole Free/Open Source software available as reusable code for just about any kind of project and think about software as some sort of Swiss Army knife.
But, the truth of the matter is, the usage patterns of a gamer are completely different from any other type of user, either from a technological and/or psychological perspective. We even tend to think of games as content in the same way as audio or video, when in fact, games are very demanding applications. That's not necessarily a bad thing, but the usability of games, their GUIs, the APIs and hardware support are not a priority and you'll see just about any of the so-called "games distro" using mostly the same software as a regular one, complete with KDE, GNOME and whatnot.
There should be only a handful of games-oriented distros, made with forks of every relevant component, but tailored exclusively for the needs of games and include no non-games related software inside. X, OpenGL, SDL and other libraries and APIs, Hardware Detection & Driver Support may seem obvious to have, but why do we need whole collections of shells, fonts, window managers or even locales? Why even the same init and authentication processes as desktop-oriented distros? Most games need to have their own, custom support for these things anyway, so the unnecessary, duplicate stuff should be removed.
Small, specialized software is better in many ways, so that the focus can be on the hardware support and the robustness of needed engines, APIs and libraries. Only then a games developer can maximize resources and focus on solving games' bugs during beta testing, and spend less time on issues with other unrelated, bloated components.
A tiny, modular LiveCD distribution is ideal for games because software diversity and versioning is better controlled, but should not be mandatory, and because the OS components can be under a free license, software houses can launch their products with the same codebase without any problem and make them either bootable or installable. Hell, some can even make professional SDKs out of it and license it to other developers.
Simply put, making a desktop-oriented distro, then just adding some drivers and some games and claiming it's a "games distro", doesn't take advantage of the technical superiority the free software community and, as a gamer, doesn't make it attractive to me, as in every distro there's some learning curve and fine tuning involved. "Damn! I just want to play a friggin' game!"
<RANT>It's a shame we're not showing of any real world usability advantages over videogame consoles or Windows-based games.</RANT>
- Otaku no naka no otaku, otaking da!!!
OH THE SHAME I fell off the wagon and use sigs again!
... what else are game developers missing?
What is missing? The Linux gamers are missing. Now calm down everyone, this is a serious point. The Linux game market is not the number of Linux users who would buy a Linux based game. That is too simplistic. The real Linux game market is the number of Linux users who would buy a Linux based game and would never buy the Windows version, would never dual boot or emulate.
The fact is that Linux users who dual boot or emulate are already customers. The developer has no financial incentive to do a Linux version, it would not generate any new money with these users. It would merely replace a Windows sale with a Linux sale. This does not rule out developers doing Linux games for non-financial reasons, like id.
When so called "experts" discuss the future of Linux gaming, speak only of the number of Linux desktops and ignore the dual boot/emulation issue, they have lost some credibility IMHO.
The Linux and Mac situations are not comparable. Linux gamers can dual boot or emulate a Win32 game. They are therefore mostly already customers of the developer. Only a handful will go without a hot game trying to hold out for a native Linux port. Mac gamers can not effectively emulate, unlike their Linux brethren they have to emulate the CPU not simply the APIs. Win32 emulation on the Mac works for business apps but not for games.
You wouldn't need to install it. That's the whole point. Game developers could create a bootable liveCD distribution specifically tailored to their game, which you'd stick in your CD drive, reboot, and it would load the kernel, drivers for your sound and video cards, the components of X that your game needs, and then launch the game.
Effectively, your PC would act just as a game console. Stick in the CD, turn it on, and you're running the game. Only difference is, if you turn it on without a CD in, you've got a general purpose computer, rather than a screen that says "NO DISC".
"City hall" in German is "Rathaus" Kinda explains a few things......
the approach is wrong.. as a gamer, my primary concern is not the OS on which the game will run, but the game itself and wether it will run on the platforms available to me... it irks me when I see a game I like and cannot play it because its for a different console or a different OS. Going back in time... the shift from DOS to Win95 was slow because most games - doom, duke3d,descent, shadow warrior etc etc ran on DOS.. and wingames was a synonym for cheap squiggly graphics puzzle oriented games. Hence DirectX was born. Loki tried to re-engineer games to be linux-compatible.. but i think that's a waste of time and resources. The designers/publishers should have a small porting team which ports the code as the game is being developed (or just after). :(
Ofcourse, Linux can be worked on to make it a a stable gaming platform - but the way its being portrayed.. its like they want it to be THE gaming platform.. a replacement.... which means the enterprise software will run on one OS and the games on another
|/________
|\A|ALYS|
Anybody else catch the gripes from one of the developers about the quality of Visual C++ code as compared to the quality of GCC code? He seemed to think that the VC++ code was better optimized, and in general regarded Windows as the better development platform. I'm not a serious programmer (I just play on on Slashdot), and my projects are small and none too complicated. Can anyone else comment on this? Is he talking sense, or blowing smoke?
Linux has a LONG way to go before it will convert someone like me. ...not that i'd ever want to), so I consider it pretty safe. No viruses or other crap, which is nice. But on the whole I don't really like Microsoft. And I don't really like being forced to use their standards and software. Switching to something else would be wonderful! ...as long as it worked.
//Begin Idiot Gamer Mode//
Who am I? A fairly typical or above average Gamer/Windows power user, i'd say. Probably above average, considering I built my computer from scratch (yay!), and recognize a handful of Linux buzzwords. Anyways, there are generally four things I use my computer for:
-Games
-Teh intarnet
-Art (PS/PSP, Maya/Max)
-Music (omgomgomg, MP3s!)
I run my quiet little Windows XP(home) box. It has plenty of the usual bandaid programs on it (Kerio/AVG/AdAware), and I try to stay away from M$ programs as much as possible (IE is only for emergencies, and I buried OE somewhere so deep and dark i'm not sure I could find it again.
So I guess that pretty much puts me dead center in the "games 4 linux" crosshairs. In theory, I should be a pretty simple convert, right? Err, actually... actually, I'm actually very resistant to Linux (please don't stone me untill after my speech, kthnx). Why? Well, lets take a trip through stupid gamer land:
Starting off with Linux in general...
-Linux? Thats that confusing OS, right? Sorry, don't have time to hunt for packages/screw with command lines/read a million help files/troll forums for answers to stupid questions. Especially not asking for help. I just know i'll get told 'RTFM' when i'm having a problem... *sigh*. If only Linux was more user friendly! Whats a rm -fr / anyways?
-Distro? Oh, gee... I don't know. There are so many! Knoppix is just for peeking. RedHat and Mandrake... aren't those "newbie" distros? I don't want to be called a newbie, so no thanks. Gentoo? Thats like, REALLY hard, right? Debian sounds fun, but I don't think i'm that smart. SuSE? Isn't that for businesses and stuff? Oh, and that Slackwhatever sounds like, impossible. Lycorsis and Lindows... pfft, I want to get AWAY from Windows, thanks. Xandros? Whats that?
Wow.. there are so many choices! None of them seem like they're targeted at ME though. And anyways... why so many? I don't want to have to choose... what if I miss out on something! Some feature that distro X has that my distro Y doesn't but I really really want? Man, i'm really frustrated and confused right now. At least with Windows its all the same...
-My hardware... um, will it all work? Drivers for my Radeon 9700 Pro? Its a GREAT gaming card... I spent a lot of money on it too. No drivers, no deal. Oh, and are there audio drivers for my sound (nForce Soundstorm) too? Ah yes, and the last thing... my entire harddrive (almost full) is NTFS. I don't want to loose 70gb of information just to use Linux! Oh, and whats all this stuff about USB and plug and play? Shouldn't that just, like, work?
-My software. Ack! I have so much of this! Lets see... I need web utilities. Already got Firefox and Thunderbird, so thats good. I'll need an FTP proggy too (I use smartFTP right now), oh, and of course, Kazaa. Some benchmarking and utility programs would be nice too (I AM a gamer after all). Soo, like Sandra, Prime, cpuz, FRAPS, etc. Oh, and I need all my pretty desktop customization programs (or equivalent) to make things look like I want... ObjectBar, Sysmetrix, Rainlendar, and LogonStudio is what I run ATM. Then i'll need media stuff... I like Sonique, and i'm trying to get more skill with Photoshop (big one), Paint Shop Pro, Maya, and Max. Oh oh, and i'll need Nero or something to burn CDs with. Ok, now onto games... yes, lots of games. I have a *ton* of classics. Everything from System Shock to Scorched Earth. They barely run under Windows though... I doubt they have Linux equivalents, though maybe WineX can figure them out? Old games can
Loki died, not because the people talking about buying Linux games were lying, but for other reasons.
Loki took on the porting or support of 21 different titles at a tune of at least $20-50k per title and royalties proportionate to if someone was selling an actual Windows game.
Loki went about the process of doing the actual publishing of the games in a manner that one would expect of a Windows publisher- thereby making the break-even levels nigh impossible to achieve.
Loki went about doing incredible, amazingly stupid things like ordering 50k units of CD's and those little metal tins for Q3:A, causing a delay in the ship date, creating impossible margins on the product when they should have ordered about 5k of the CD's and used DVD boxes to cut costs and get the official Linux version in people's hands in about the same timeframe as the official release (So that people wouldn't have went and bought the Windows version and "patched" it with the binaries set from Id...).
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
And I don't give a rat's arse if they're open source. I want them fast, I want them prominently available from hardware vendors and/or distributors, and most of all, I don't want to have to play a Towers of Hannoi with dependencies and command lines to get them to install and work on a stock Red Hat or SuSE system.
Oh, I know it should be simple enough, but it isn't. Google for problems with (e.g.) NVidia drivers with SuSE distros and that should give you a sample of the fun that awaits. For every twitchy zealot who'll chime in saying "Well, it just worked on my system!" (even though Linux cognitive dissonance means it probably didn't "just work") there will be someone who eventually got it to work after hours of hacking and begging for help in forums, someone who gave up on it, and someone who thinks it worked but who is still using old drivers without knowing it because they missed the "Wrong version of fleem" error in the forty screens of script output that ended with an "Install complete."
If you were blocking sigs, you wouldn't have to read this.