Hi Ive read your article and it's spot on!
Games are the only reason I'm still using Windows and I know that if Counterstrike and Battlefield2 could run natively on a linux distro, I'd drop Windows tomorrow.

I work in IT, all grown up and all but I still like to play computergames often and it's frustrating not being able to use a platform of my choice.

Of course I could try and tweak a Linux installation and make games run well on it but that's time and effort I'm not prepared to spend when I can play games effortlessly on a Windows machine.

For all intents and purposes Linux would suit my needs better than Windows xcept for that one crucial thing.

Maybe the Linux advocates sofar have been too intelligent or geeky to see that the biggest problem facing desktop linux at the moment is nothing more than a lack of the bigger game titles and an easy way of setting up a linux system for gaming purposes (I'm primarily an FPS PC-gamer so thats ofcourse the area I'm interested in).

The key is ease of use. They've come a long way with Ubuntu and Mandriva and others so I'm sure it can be done.

I wish you the best of luck with your endeavour and keep preaching your message because it's been neglected for far too long now.

NO NO NO! What Linux really needs is for more developers to concentrate on SDL or something similar. What Linux needs are a GREAT (not just "good enough") set of APIs (like DirectX) that will be competitive with its rival. If there was a complete package (like DirectX) that provided video, audio, networking, input and whatever else is there (and no SDL isn't mature enough to compete directly with DirectX) then it would attract developers. If there was a Open Source package like that, then more developers could possibily consider programming with that. Say if it was available on not just linux but also Macs and Windows, then perhaps developers in the future will consider using that set of APIs. And by doing that, the games could be available on all platforms.

Developers don't want to spend time porting from one platform to another, specially when the other platform is difficult to work with and the market being so small.

However if there was a cross platform set of API's (packaged like DirectX), then it would appeal to them since the porting would now be very very minimal.

I'm not a programmer (although I am learning java) but that's just my random thoughts. I could be totally (probably am) but that's just my 2 cents =]

'The main problem with Linux at the moment,' writes Ian Bonham, 'is that to get Linux mainstream and to build a new generation of writers, we need a killer app - we need a decent game. We need a game in the F/OSS that everyone wants, everyone can see how it works, and learn from it.'

I see so much great interaction and ideas and people saying that they're willing to help. I guess some of us, if we really are that interested, should stop being sayers and start being doers.

I would like to get a hold of a developer who could make a simple trivia game for ps2

We can't dictate to the game publishers.

Before the IBM PC came to the market there were thousands of manufacturers of various components that would work in one hardware environment or another. The market was nonstandardized and fragmented. The PC changed that.

Before windows, the DOS systems were nonstandardized and fragmented. Game publishers had to include many drivers and hooks for others that might be needed later. Windows changed that.

The Linux world is still nonstandardized and fragmented with dozens of distros each with unique installation features. Games are written for Windows because Microsoft does 90% of the work for the game's programmers. Before Linux can be a serious challenge to Windows we have to have the same type of infrastructure that the Windows registry and DirectX offer to the publishers.

I don't agree on one spacific "killer game", as this is a question of taste and would only be a momentaneous solution for a few linux enthusiasts that wouldn't have any big future or achieve a change within the gaming market. What needs to be done is to adress as many game developing companies as possible to program their games in openGL only and not in direct3d anymore and offer their products to be able to run out of the box on a linux machine, with installer included. As I posted before, alo the 5.1 issue needs to be adressed.
Another possibility is to achieve, that companies licencing engines (like id or epic) do so obliging the licencees to include linux installers/openGL for their games based on the licenced engine. That way the thing could spread more and more, and maybe in a mid or long term make openGL-games even more popular than direct3d games and bring along a new criteria for what quality in games is.

But the question is, why game developers should do this, as they program their games in direct3d because they get support from Microsoft. To push them in the openGL direction, there would need to be another powerful commercial project behind it, that also gives support. It could and should be a strategic alliance of several software and hardware companies, that don't like Microsoft for any reason (maybe even IBM, Sun Microsystems...). Only that way I see a future for implementing openGL in most games in the future, or would you think it a better idea to licencing direct3d-Technology from Microsoft, in order to implement it in Linux/Unix to be able to play games? And in the end, we shouldn't forget that openGL was there before direct3d and it was an API meeting professional
standards, while the first direct3d games were crap, running in 640x480 pixels and 16 bit with poor effects only.

That Killer game that the Linux platform needs is named World of Warcraft(MMORPG).

An e-petition is being started invoking all Linux users to request VUGames and Blizzard Entertainment to port World of Warcraft to Linux. Blizzard has been known in the past for their support to PC and Mac users with hybrid games. The aim of the petition is to have a PC/Mac OS X/Linux hybrid. Sam Lantinga, developer of the SDL API and ex-employee of Loki, is known to be Software Engineer of World of Warcraft since 2001. Among his duties in Loki was to port Mac games to Linux.

World of Warcraft fans are hoping that the petition may encourage VUGames to fund the project. Unreal Tournament, Quake 3, Neverwinter Nights, Return to Castle Wolfenstein and Wolfenstein: Enemy Territory are among the games that have being ported to Linux by other companies. Knowing the wide-reputation of Blizzard Entertainment, it would be a boost to the Linux community to have a MMORPG of this calibre. NVidia has Linux and FreeBSD drivers, which confirms that the Linux platform is becoming fastly a Gaming market.

The petition can be signed Here
http://www.blizzpub.net/petition/

We recommend Linux users to spread the word. If you want to help, contact me. I am sure you know how important this MMORPG is for the Linux community before the Gaming Industry.

Well, I don't agree totally. The only real reason I keep Windows at home is because I can't find any good Video Editing package for Linux. On Windows there is a plethora to choose from (I use ULead VideoStudio 7).

yeah, absolutely right, because the only reason, why people still need windows is because of the games. But also other problems with linux need to be adressed, e.g. 5.1-sourround sound in games, it is not just a graphical issue.

amd64 + linux should do the trick to make a killer game in 64 bit area

I believe we ARE getting killer games. We're getting Unreal and Quake. The main thing is that they're not Linux-Only, so Windows users don't really have any incentive to drop Windows and use Linux when they can use Windows to play the same game. If I was a developer, I would definately never drop support for Windows until I saw that I was beginning to lose revenue from that audience. We also have Return to Castle Wolfenstein(although it isn't relly the most popular thing out there now, I do remember having a lot of fun with that thing on single-player mode). The only thing really about Wolfenstein is that they don't MARKET the Linux version. From what I've read(And I've only played it on XP), is that people have to download a demo for Linux, then copy over the .wad file on their disk, or something like that. I'm not really sure, but I do know they support Linux because they have a Linux patch listed, and it talks about single player fixes and the like.

http://www.idsoftware.com/games/wolfenstein/rtcw/index.php?game_section=...

and here's a demo for UnrealTourny 2004 for Linux

http://www.unrealtournament.com/ut2004/downloads.php

The idea of using a live CD for games sounds good to me, but I am a linux newbie with no understanding of game development, programming, or anything else that could even be considered technical. So I have some questions.

1. Would it be possible to launch without restarting. When I am sitting at my computer doing homework and I get bored would I be able to start a game and play for twenty minutes then go back to work without having to restart twice.

2. If the liveCD was soley to be used on the one game would it need a GUI or any comands at all, could it just launch the game. Also what type of performance loss would there be.

3. Could a liveCD be used as a way to make games ,and everything else, run on any platform. Perhaps if the liveCD would be able to run on both Mac and linux gaming companys would produce it themselves.

4. Would computer gaming lose any of its coolness. One of the things that make Computer games different than consoles is the ability to save so much information to the harddrive, enabling more customization in charachters and maps. Sometimes you can play games without messing with the CD.
Also could updates and expansions be played.

I think the Stratagus engine should be included in anything like this... because it has some great datasets already, and it's able to allow people to create new RTS games... :)

Ratchet

Not sure why, but the second quote I made didn't go through. So here it is...

Quote from Steve Baker (above), making an excellent point:

"Furthermore, most game art is terribly under-utilised. Look at the 'Crazy Taxi' classic (it's one of my favorites). It contains a 3D model of a MASSIVE chunk of San Francisco.

It's very detailed. It was used (afaict) on just one game.

If someone released a 'taxi' game with that data set - then someone else could pick it up and produce (say) a skate-boarding game using the same city with the addition of a few ramps and grinding rails. Someone else could turn down the lighting - render it in monochrome and produce a moody detective story/mystery - someone else can add spaceships and alien tripods and produce an "Aliens invade downtown San-Francisco" game...we'd get a half dozen reasonable games from one really good data set."

Thanks,

-Beetlejuice

K, you guys need to stop talking in circles...I am getting dizzy!!! Can't make a great game cause there's no programmers...but then theres not enough artists...but then theres not enough money...but...but...but... When did "but's" stop anyone in the Open Source community? Users or developers? We all contribute to Linux's success!

I'll tell you what, I think that "Steve Baker" (see above) knows what he's talking about. Linux "really" lacks in the artist category (see below).

>The trouble with the OpenSource world is that we can only
>head in the diametrically opposite direction: What can we
>do with a LOT of software and hardly any 3D models? The
>answer is precisely the heap of mediocre games you see on
>HappyPenguin every day. (My games are no better - I'm
>ashamed to say).

He also makes a good point here!...

> The issues of art it seems will make or break opensource
> gaming so that solving the art issue in an open source
> way could be the killer game that linux needs.

I have no hesitation in saying that there are PLENTY of great game ideas - and more than enough enthusiastic and talented programmers to produce any game concept you could come up with. If you can solve the art problem then I see no reason why we couldn't write commercial quality games as OpenSourced products.

Furthermore, most game art is terribly under-utilised. Look at the 'Crazy Taxi' classic (it's one of my favorites). It contains a 3D model of a MASSIVE chunk of San Francisco.

It's very detailed. It was used (afaict) on just one game.

If someone released a 'taxi' game with that data set - then someone else could pick it up and produce (say) a skate-boarding game using the same city with the addition of a few ramps and grinding rails. Someone else could turn down the lighting - render it in monochrome and produce a moody detective story/mystery - someone else can add spaceships and alien tripods and produce an "Aliens invade downtown San-Francisco" game...we'd get a half dozen reasonable games from one really good data set.

> Since you seem to really know what is needed you could
> probably draw up a sepec doc for a gameart project which
> will fill such a need.

That's something that would depend on the game that was being produced...and because everyone is working for free, the game concept has to be something that your chosen artists would enjoy working on.

So - IMHO - the right approach would be to first find a bunch of enthusiastic 'OpenArt' guys with 3D modelling skills. Let THEM pick the genre and look of the game - then put out a call for help for programmers to come and make it work. I guarantee that any halfway reasonable artwork team could attract any number of enthusiastic programmers to write the code.

> There are tons of wippersnapper teenagers who are graphic
> wizzards who might be coaxed into joining.

I beg to differ. Numerous projects have failed due to lack of artwork. Those teenagers (if indeed they exist) either can't or won't do what's needed.

> And if the big talkers who started this thread would add
> some funding for such a project who knows....

Well - that's a different matter. With a big pile of money, you could indeed get some paid artists on board. But it's not cheap. Whoever paid would probably need a return on their investment...then the programmers would get pissed because they AREN'T being paid and this is going to be a commercial project...then it *IS* a full blown commercial project with all the commercial demands that mean that this has to be a Windows game and not a Linux game.

I don't see that as a solution - if it was, commercial companies would already be writing Linux-first games.

(whistlingtony)
"A game does not have to have the flashiest graphics to be successful. Look at Half Life! It's mods are STILL played heavily today. All it takes is a GOOD GAME, good playability. we'll forgive blocky 1999 graphics for a good storyline and kick ass internet gameplay."

A good game is essential, I agree with you there 100%, but it is not everything. Remember that people do not just play games simply for gameplay, they also sometimes, to varying degrees, play them for sensory distraction (read: tuning out the rest of the world). Go to a LAN party sometime and watch everybody get sucked into their computers in a massive, dark room. The more immersive a game is, the more people get sucked into it. The more that happens, the better the response to the game. Good responses generate word-of-mouth advertising, which perpetuates the product and brings in new "customers".

This brings up an interesting issue - there is not much of motivation for experienced professionals to make a really, really great game for Linux because nobody will have any reason to pay for it if you just give it away for free. Nowadays, getting a pirated copy of a game usually involves some kind of workaround, but is not impossible most of the time, so actually buying a game is more of a convenience/legal thing than one of necessity. "Selling" a F/OSS game in places like Best Buy and Electronics Boutique where there's some kind of "support"? Forget that. If you need technical support to run a video game, you've bought the wrong video game.

What is needed is *some* kind of transactional mechanism whyby a developer can sustain themselves (i.e. eat, have a place to live, etc.) while working full-time on such projects. And I'm sure if that could be done, it would lead to many more good things for Linux as other kinds of software could follow suit.

I will say this: most of the F/OSS games that I've played that were made specifically for F/OSS (read: not a linux port, but a real linux game) are somewhat lacking in the content department. The code might be wonderful, but if the game looks like $#!% and there is something with comparable gameplay that looks better, which one do you think people are going to play?

Also, if Linux even wants a hope of competing in a market dominated by consoles (even in the off years like this one), there must be excellent gamepad/joystick support. The people suggesting an open/common hardware library are right on point. When I sit down and think about a game design, one of the first things I do is get out a gamepad and try to work out a simple input scheme that will allow the player to do all the cool things they should be able to do in the game. Why do this instead of keyboard mapping? Gamepads, surpise surprise, are designed to serve as comfortable input interfaces for games. If your gamepad support is even remotely lacking, you are stiffing the game community, effectively limiting yourself to keyboard/mouse combos, which does give you mouselook and graphical screen selection, but kills anything that doesn't follow that specific input paradigm. So you're stuck with mostly FPS and RTS style games, leaving out racing, fighting, non-FPS action, and other genres of gameplay. PCs have an advantage over consoles in that many different kinds of input devices are available for people to buy. The problem is somebody needs to write drivers that work for all of those devices.

I really hope that the Linux community addresses these issues because it is difficult for me to justify spending my time developing a game for Linux right now. If it were a more feasible thing to do, there is no doubt that I would push for it. And if were good enough for me, I'm sure it would be good enough for enough people to really make Linux into the mainstream home desktop market.

There is a chance that the MMORPG Saga of Ryzom (http://www.ryzom.com, currently in closed beta) will be ported to linux. It uses the NeL (http://www.nevrax.org) engine which is under GPL which indicates that the final game maybe will be under the GPL, too (it should, IMHO). SoR can compete with contemporary MMORPGs. Artwork will be copyrighted and you will need to pay a subscription to play on the Nevrax server, but at least we would have a nice MMORPG that runs (natively) on linux.

The issues of art it seems will make or break opensource gaming so that solving the art issue in an open source way could be the killer game that linux needs.

Since you seem to really know what is needed you could probably draw up a sepec doc for a gameart project which will fill such a need.

There are tons of wippersnapper teenagers who are graphic wizzards who might be coaxed into joining.

And if the big talkers who started this thread would add some funding for such a project who knows....

Aaron

Oh - and the other problem...the artists who give stuff away for free want to design the 'hero' pieces. The really glamorous sports cars - the beautiful and artistically interesting pieces. You don't see them designing free trash can models. 99% of game art is the filler stuff.

Also, game art needs to pay careful attention to things like texture map consumption and polygon counts. Turbosquid is fill of very beautiful 100,000 triangle car models. But for a game you'd maybe want a 300 triangle model that perhaps doesn't look *quite* as good - but which we can draw dozens of and still make a decent frame rate.

C++??? Let's hope not... Real ANSI C, sure.

And, a "game based" distro? Hmmm.. No.

The "When will Linux be a mainstream desktop OS" debate has been going for at least as long as Linux itself has been around. Want to see what it will take for Linux to gain mainstream acceptance?? Spend some time using Mac. OS X. That is how Desktop Unix should be done.

As for trying to emulate Nintendo and Playstation? Let's leave that to Microsoft... There's a market they can keep!!!

In reply to the art repository.

I understand what you mean thats why I meant specifically for games.

To create a artgame platform which spells out art that can be used and reused for games.

I wasn't thinking really about a respostory more likely a platform or a concept which will make creating art for games easier and empahsize reusability.

this is not meant to grab helter shelter art. The same is true of music, chances are good that music grabbed from a repository won't work with a game
Aaron

There already are repositories for free 3D artwork - check out all the free stuff on TurboSquid for example. However, that doesn't help.

A music library makes sense - because you can pick up a music file and you have music for a complete game level perhaps.

For art, it's useless to have one person design a space ship - another person design a house - another a car. To make a game, you need a complete level. If you have a thousand buildings, they can't all have unique textures - they have to fit together, they have to look visually like they belong together. If you put a house with really nice hand-painted textures next to another one with photo's of a real house pasted onto the walls - then even though each one individually looks good - they don't look right together.

It's really difficult to do anything useful with art that wasn't designed specifically for the game.

Hmn,
If artwork is the problem, a linux solution should be the answer.

Libragameart.

I just made it up.
Libramusic is free music available to all. So game makers should post their existant art and make a framework so that game art can be reused.

The detail would need to be worked out but this could be a nice beginning.
Maybe it exists already?

Aaron

I'm a long-time amateur games writer. Tux-the-Penguin, A Quest for Herring was the first 3D game for Linux - and my PLIB library suite has gotten quite a few other major games started (eg Torcs and FlightGear). So I've been here a while.

The problem (and it truly seems insurmountable) is that there are PLENTY of people with good ideas for games. There are fewer (but still enough) programmers with the skills needed to write the code. What we are totally and utterly missing is artists...artists who'll work for nothing ...artists with games skills.

I signed up for the LGP effort to start up a games company ('AngryPixels'). Same deal - lots of great ideas - lots of wildly enthusiastic programmers just chafing at the bit to write something great...but ultimately, not a single decent artist.

I'd *REALLY* like this effort to make a Linux games distro happen - and I'm willing to help out in every way that's reasonable...but unless there is a good answer to the artist gap, it simply won't happen.

Think of a car driving game for example. I was able to use my libraries - along with the ODE physics library - to get a basic driving game together in about three weekends. However, to design a SINGLE roadside building takes an entire evening. Multiply that up to the size of the Crazy Taxi city - or the GTA city - and you have over a thousand buildings to make - not to mention lighting poles, phone booths, dumpsters, trash cans, trees, flower beds, signs..... If I could stand to do this, it would take me maybe 100 times as long to make the buildings as it took me to write the code...and that's without thinking about the artwork for cars, special effects, etc.

The trend for modern games is to increase dramatically the number and quality of 3D models to create the depth and variety - and to write a very generalised, data-driven engine to draw it, control physics and AI. Most commercial games companies have ten times as many artists as programmers.

The trouble with the OpenSource world is that we can only head in the diametrically opposite direction: What can we do with a LOT of software and hardly any 3D models? The answer is precisely the heap of mediocre games you see on HappyPenguin every day. (My games are no better - I'm ashamed to say).

So - if there a good solution to this, I'll be on the team in a heartbeat...but I'm not optimistic.

Usability is the main criteria in popularity when we talking about end user (not specialists with specific knowledge). People want to receive more stuff for money.
Applications (not only games) are the fuel for popularity of any OS or computer platform. And this is not a secret that is quite expensive to develop these applications for different platforms at the same time. And there is no way to convince commercial software manufacturers to support platform that is not very popular among ordinary users. And this cannot be solved with one or several killer apps. People will not install unknown OS (or buy unknown computer) only because these particular apps.
No matter how great this platform is technically. As soon it is not popular among end user it is doomed to be niche product.
So I think that as soon as it is not such easy to create original popularity for Linux among end user, why don’t “borrow” some popularity from known platforms: M$ Windows, PS2, Xbox, Nintendo and etc.
It is not such a bad idea to put emulator for known platforms under Linux. And if they would work decently it could be good start for popularization of the Linux as end user platform. But they (emulators) should work at least not worse and cost less then native platforms.
I foresee some critics from “pure play” supporters: Like, it is not the way of glory to emulate commercial platforms. But I do not say to stop development of native applications at all. Let them proceed and hope that they could produce something native with commercial quality tomorrow. But for a while there will be a lot of end user applications with commercial quality and end user will not have to have dual-boot configuration.

People, I have become involved with a project that aims to do just this (well, we are still debating whether to sell copies as well as allow it to be downloaded). Our website is http://highrisk-game.sourceforge.net so come and visit us.

Oh, and on the forums, if you get that far, and wih to identify me, I am MorphiusFaydal.

I am not really a gamer, and if I do get time to play games, it's on a PS2. However, I believe that there is a large market of potential Linux gamers out there waiting to be tapped. Many of these people use Linux for serious
work, and Windows for games. However, how many (primarily) Windows users use Linux for anything that Windows cannot do?

Major software houses like ID would have to make a loss-leading investment in the Linux gaming market in order to create the momentum, by porting their major titles directly to the Linux platform (DirectFB maybe?), and NOT through the likes of Wine(X), unless they can find some way of eliminating the performance hit.

A good way to approach this might be to base games on a self-booting and skinny Linux distro, along the lines of eMovix. This approach (might) mean that the same game could work on an X-Box and on an Intel workstation (it may need some system detection code in the bootloader). It should also be made possible to run the game on a workstation that is already running Linux - it should automatically install any extra software it needs.

Obviously for commercial games, putting the source code on the CD is out of the question, but it would be very helpful if software houses would help seed the market by providing guides and tutorials, sharing their experiences and GPL'ing code from the games that they give away free on magazine cover discs.

Opinions?

Rohan Beckles
rohan.beckles@virgin.net

Re:Shamkao commented on 15 March 2004:
I think it would be a good idea to put some of this stuff:
- Emulators of clasics computers
- Mame emulator <-- There is one http://www.distrowatch.com/index.php?distribution=knoppixmame&month=all&...

- A xbox emulator
- A PS2 emulator
- Games like FreeCiv, FreeCraft, Quake ...

What I don't understand is why when people think about Linux games, they use a totally different logic than when thinking about other applications.

Take for example

> The "standard" for a killer game is so high these days that the production cost is out of reach for most developers.

> The other huge problem is everyone wants something, but few have the time or energy to do it.

> Until linux users are will[ing] to pay for software there will not be games.

> People are not going to pay to play in linux.

Substitute the word "game(s)" in these statements with any other type of application, and you get falsehoods. All of them boil down to Bill Gates' 1976 question in his open letter to hobbyists:

"Who can afford to do professional work for nothing? What hobbyist can put 3-man years into programming, finding all bugs, documenting his product and distribute for free?"

I would have hoped by now that Linux and the whole FOSS movement have given a pretty clear answer: There are plenty of volunteer developers all around the world willing to contribute to software projects which are distributed for free.

Nobody whinges about the fact that Microsoft haven't released Internet Explorer or Outlook Express for Linux, or that there's no native Linux port of Adobe Photoshop, do they? Because despite the huge development costs of these applications, there are free alternatives available for Linux which are as good, or better, than their commercial equivalents.

So why can't the same be true for games? Why spend so much time and effort trying to convince commercial game development companies to release Linux versions, when we could be writing our own high-quality games that are just as much fun to play?

[For anyone who thinks it's too hard, I recommend you check out the Scratchware Manifesto (http://www.the-underdogs.org/scratch.php) which examines the common assumptions about what's required to make a great game.]

Let's encourage people to try writing their own games on Linux, on their own or in collaboration with others. As long as the available graphics APIs are powerful and easy to use, we have all we need to start a Linux gaming renaissance. (If somebody ambitious wants to try and implement a free implementation of DirectX, like Mono is for .NET, that would be a good way of helping developers cross over, but it's not essential.)

Go on, let's show them what we are capable of.

We Need Games? Well then you need to search a little more.
How about the new Unreal 2004 Tournament? How about Return To Castel Wolfenstien Enemy Terrority? How about www.racer.nl for a Need For Speed type of game?

What we do need is games Like EA's Need for Speed, Maden Football, Baseball, We need games like the ones Disney put's out for Kids. We need a good flight Simulator, though http://www.flightgear.org looks good. Or Nascar RAcing and Battle Field 1942 would be good. Cube does look good but I think we could do better.

Why not start a patition to these companies to see if they can get off their cans and help a new movement in gamming?

Bonham do you know CUBE?

It's a 3d good game (open obviously):
http://wouter.fov120.com/cube/

well, if there was such a thing I'd immediatelly get down to downloading it and spreading it around to at least 10 people :P

Don't think Gaming distro, think gaming platform. Give developers a) a common platorm to develop for, and b) some good development tools. Think in terms of turning every PC in the world into a gaming console... Give developers an easy way to add their native game to small live distro which can then be booted directly from CD/DVD. If it's cost effective to develop games for Playstations, Game Cubes, and Xbox's, it's got to be cost effective to develop games for every PC around.

The easy solution may be to have WineX included in majors Linux distros.

We could wait years before seeing Linux native major games, so right now we should live with DirectX and there is WineX for such purposes.

Please Mr Transgaming, Suse, Redhat, Debian, find some agreements to see WineX as a standard package in modern Linux distros.

What we need is an excelent cross platform Direct X replacement library.
It can interact with real directX drivers on Windows and interact with OpenGL sound and such like on other platforms.

Then we can persuade developers to support it instead of DirectX, and if it is similar to DirectX it wont be too hard to get them to migrate.
So long as the API is common on all platforms and there is no performance hit it will go down a treat.

Gaming under linux sounds cool, we just really need some better DirectX support and audio drivers because opengl games [quake,jedi outcast/academy] works here without problem. Only think that is pain in the ass is audio, sound is just not that great as at windows. I migrated recently ( after an 1/2 year i was able to turn my back on windows ). Linux gaming is defenatly best solution for the user problem because many users dont really know whats happening, they just click the button or run the program but dont know how it works. Some kids want to learn but you have to show them that sometimes its cool to learn such things as programing.

This idea is great and i think you should get it working.

>>> Loenner:
What Linux really need is
1) a common standard for program installs
This should somehow also solve the package dependant hell, that all the known package managers have today
>>>

>>> Dan Groom:
The way around dependecy hell is to assume most users won't have linux installed and just to run a standardised distro off a bootable CD.
>>>

Live CDs are a great way to show off capability, but the several-minute delay while you wait for the system to boot will put people off. (Yes, live CDs should still be used, but perhaps not as a primary method.)

What about moving the OS boot process onto the hard drive, while still loading all the application software from CD? This could be done with a fast-booting mini distribution optimised purely for this purpose. This is a plan that some friends and I have actually planned to implement at some point:

- The OS itself contains exactly what is needed to boot up ready to load applications, no more. Even then, only the currently essential services are started up front, to improve boot time. Only the most common (and least evolving) shared libraries, beyond essentials such as video and sound, are included here.

- Games are repackaged in a special form intended for burning to CD. Each package contains all of the non-core dependencies of the game in the standard filesystem layout. (Yes, this will bloat the package size, but remember each game is intended for its own CD.) It also specifies any required services (eg, network), so the OS can start them before the game is launched.

- When a CD is inserted, it automounts and is checked for packages. If a package is found, the package filesystem is mounted "over the top" of the OS filesystem, and the game is launched. (If multiple packages are stored on a CD, a menu is presented first.)

Going one step further, each game should be modified (if necessary) to support play via USB game controllers. You've now got yourself a Linux console system.

I am now witnessing a flow of classic dos games for window, on Linux, freedos, and dosbox.

I see that many older games can now be played on Linux.

but what I don't see is some vision unique to linux.

let me explain please. Windows professional audio has a huge user base and until recently there was little to no interest in Linux for audio (MAC still rules here) but when a group of audio hackers pushed for lowlatency and a unique server client system (jackd) and a few killer apps and presto.

With games on linux to work, a group of developers need to come up with what would make linux special for gaming and create the core tools to achieve this.

I might add that integreating the lowlatency and audio tools would certainly be cool.

But more is needed. If linux only copies windows than who needs windows.

I recall a hacker (17 or 18 yrs old) saying that he dual boots windows only because of games.

Making linux a plaform for game development is a real challenge.

Aaron

What's NEEDED is an OS that installs without anything more than the most basic of questions, with the options that home users need most. Add to that an easy to use GUI with a unified method of adding applications that NEVER EVER required anything more than doubleclicking and hitting next over and over again. Also, developing applications should be easily done in an Object Oriented environment so that just about anyone could plop together small, internal appliccations for their businesses to use. Also, verify that the majority of your users will never have to see a command line for ANYTHING they do unless there's a problem and a tech on the phone. And then make sure that hardware and software can be easily removed, reinstalled, detected, switched, rearranged, and managed by even the most mediocre of home users. Then, once you have something that looks kind of like Windows, fix the BAD stuff, like security, add a couple of whiz bang features, and you should be good to go.

Forget it, back to loki times, in two years 10 great, veru great were out... you know how many sells did they do for their greatest hits (q3 and civ II ? 5000 each).
not more, not less.......

People are not going to pay to play in linux.. why ? because it's not easy. Running mozilla, ok.....
having a huge game running and setting up the libs for it it's a pain on the ass.......

and just to let you know, it's easy to grab some already knoppix full set mame in p2P network.. just put your cd and play......
but IMHO, why should I bother play game in linux when I got my xbox, ps2, game cube and tons of great games for almost free (the p2p, always it.......) ?

I don't agree whith this kind of dream : one machine for everything.... I want my child to play, not waiting I've finished to set up the weel scroll in zarkanoid version beta 0.2, works only when the moon is fulll.......

you noob.......

"let's learn C++ to the children", of course... and after that, let's learn them how quantic physics and voodoo.......

C++ requires advanced skills in algorithmic, in abstraction and even for informatician students it's not easy......

basic is the best langage for beginners, ever.....
C++ maybe the hardest ever......

quel rammassis de conneries.........

Holy Craft Macaroni & Cheese!

If you built a mini graphic distro of linux, the game would be esentially platform independant!

Niz, i dont want the sources on the cd.
I come from a linuix from scratch, so its not that im the kinda guy that dont use source :D
But the advantage of a package is that it is a binary and you dont have to fiddle with it. indeed i think that a distro fails if you ever have to compile. that either means that their binary isnt compiled with all the options you can get from source, or that there isnt a package (binary) for the distro.
So source, on ftp.
The gentoo games sounds good, but i guess its only a gentoo branch.
I wud go in that direction, steal gentoo's portage and totally prepare the distro for games, and emulating windows binaries. The most important front end shud be the graphics card configurator. And there shud be an export mode for all frontends so geeks like us dont get scared, too.
Support for closed source kernel graphics card modules. Wine with glide. Lotta stuff!

Right on!!!!
Been working on cell phones for 17 years, and there (as well) technology is driven by kids/teenagers.

Go 4 it!!!
Regards

Well, as the Gaming Industry editor of LinuxWorld Magazine, let me point out that there's LOTS of content about the state of gaming in Linux in our April issue. Watch for it when it hits the newsstands!

In reply to "the Dude"'s rant:
> We don't need the source on the CD, what possible use...
> The author of this article unfortunately has his head
> stuck in the sand...

Thanks for assuming you speak for all of us there *NOT*

Don't drag all of us us down to your level of the Microsoft-mindset non-technical user. Here's the scoop: Most Linux users ARE techies and DONT WANT and OS thats been dumbed down (read: castrated) for the masses.. You don't like source code with your stuff? Go back to XP. I know its incomprehensible to you "plug and play" types, but some of us actually *can* understand and do things with the source code.

Besides, just because its on the CD doesn't mean you HAVE to look at it.. or does an extra directory on the CD confuse you too much?

For those commenting on the "dependency hell" - please, realize, that Linux <> RedHat ! The problem of dependencies was solved ages ago already - Debian with apt, Mandrake uses urpmi (which works on top of rpm), Gentoo for hard-core tinkerers. That is not a problem.

Problem is elsewhere - until you are able to sell few hundred thousand boxes of the game, nobody is going to even consider porting something to Linux. Until your kids, their friends and your grandma and her neigbor are using Windows for everything, you will not get games for Linux. It is just not worth to invest time and money into a game, for which you have to install and learn different OS (even with self-bootable CDs) and when you are done with the game, you have to go back to Windows, because you do everything else with it.

The whole problem is just economy of scale. The games are very expensive to develop and nobody is going to bother for few "freaks" using a niche OS. Except of maybe some students, but then do not expect another Doom III or UT from them.

We don't need the source on the CD, what possible use do 98% of the game playing public have for it? The answer is nothing, for those not up on their coffee intake.

The author of this article unfortunately has his head stuck in the sand, you don't want to put on it what you would want, but what the average dude would want.

Its been done!
RTCW:ET
It was a killer game for a while
It was free
It was given away with a linux live CD and it was availible for free download for linux.

It didn't work.

Your wrong, linux needs lots of things.. and what we need from the gaming industry we're currenlty getting (we could do with more of it though).
What we need is more companies to release their games with either in the box linux binarys or do like they did for UT, and release linux binaries afterwards..
Most importantly we need Sierra to support linux.
Sierra has some of the best selling titles, but they have ALWAYS turned their back on linux.
batsards.