Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
Linux and Games : The Good, The Bad, and The Bankrupt

By carbon in Technology
Thu Feb 07, 2002 at 06:40:45 AM EST
Tags: Software (all tags)
Software

Open Source Software and UNIX have come a long way in the past few years. They've successfully made the jump from a server platform to a viable desktop platform, at least for those who aren't computer illiterate. The jump was characterized by the addition of a new property onto the list that the developers needed to keep track of (from stability, speed, and power to stability, speed, power, and intuitive ease of use), and despite all the naysayers, the KDE and GNOME projects demonstrated UNIX's usability and Open Software's suitability as a general desktop environemnt, as well as on servers and as network tools of all kinds, development workstations, and even office systems (for example, the Korean HancomOffice/Linux/KDE deal). But, there is still a major facet of computing that UNIX has not demonstrated it's ability to overcome, and that is gaming. At least, it hasn't until now...


Those who follow these things have probably become aware of several major leaps, bounds, and trip-ups recently in the world of Linux gaming. Most notably (and sadly), Loki Games has closed shop. They've ported many games to Linux and seemed to be turning a profit, but apparently they just weren't doing well enough, and have stopped selling games as of January 31st.

Although they did not succeed financially in the long term, Loki's temporary success did prove a few things about UNIX and gaming. They proved that the UNIX platform is at least capable of running many Windows games, often with equal or beter levels of performance and stability. They also demonstratred that the Linux community was at least enough into gaming for Loki or a company like Loki to keep afloat for a few years. These are two very important points, because they show that Linux could become a commercial gaming platform as or more powerful then the ubiqtuous Windows, if only more people used it.

On a brighter note, the Adonthell Project has released version 0.3 of their RPG engine, as well as a fun mini-RPG called Waste's Edge. Although they still don't have a battle engine, they've shown that they don't need one to make a game enjoyable and challenging (I still haven't been able to figure out who took the jewels!). I can't emphasize enough how impressed I am with how well this game is put together. The dialogue is well coded and eloquently written in a Tolkeinesque style, the art is both unique and cool (uniquely cool?) and the engine seems to be able to handle anything they or I throw at it. I think this is the most promising free game in development at this point. That's saying something, since though there are plenty of games which have serious potential to be the killer OSS game, suchg as WorldForge, Parsec, Pingus, and many others.

Another interesting UNIX game is Tux Racer, which has an interesting past, originating from an OpenGL exercise and slowly maturing into a full-fledged Open Source game. Since then, the original developer of the game has written a commercial version of the game, and released a demo of it. Though the new Tux Racer isn't OSS, it does run on both Linux and Windows from the same disc, and the demo is very cool, with new graphics, new animations, a new level engine, and most interestngly, a single-station multiplayer mode. It isn't perfect (the framerate isn't that great, and the game will be DOA unless there's some method of importing user created levels), but it is yet another good testament to the usability of UNIX as a gaming platform. The fork of the Open Source Tux Racer 0.69a, called Open Racer also seems to be developing smoothly.

So, as the above information hopefully shows, UNIX and Linux in particular does indeed seem to be capable of running games well. There area lot of unfinished and just plain uninteresting Linux games out there, of course, but these seem to be generally overwhelmed by the good stuff, just as in other OSS catagories. Also, I don't believe Loki failing is a sign that Linux users simply don't want to play games, as this is the single thing I hear the most requests about, and the single reason that the majority of Linux users keep a Windows partition or a copy of WineX around to play their favorite games with. The big problem, then, is not demand or technical feasability, but just getting the games developed and in the hands of the gamers. But I'm not too worried about this aspect.

Game development seems to be approaching that stage of critical mass that the UNIX desktop went through 2 or 3 years ago. For a long time, people thought Linux/BSD was and forever would be useless as a non-developer workstation, and then all of a sudden, all sorts of different projects seemed to simeltanously come into maturity. I remember using FVWM and WindowMaker in one version of RedHat and then trying KDE, GNOME, and Enlightenment in the next, and being completely floored by how all the stuff that the developers of these projects were talking about just magically seemed to come together and become usable. Loooking at games, you can see a similar evolutionary parallel (lots of development going on in the background, many unsuccessful projects with few developers and a couple successful projects with many developers, user oriented catagory), with the exception of commercial games on Linux.

My final conclusion is that both Open and commercial gaming on Linux is now just about to come to fruitation. Some projects aren't doing that well, while others are progressing with incredible speed, but this is a normal aspect of the evolutionary development process. The new property added for the developers to keep track of in the jump from desktops to games development is gameplay, but from the looks of things, this hasn't turned out to be a significant probelm to the resourceful OSS developers. I think that this final major hurdle keeping LInux off the desktops of the masses is finally being overcome, and I'm very excited about the future of Open Source gaming.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
Is UNIX a good gaming platform?
o No, it's fundamantelly unsuitable for gaming. 17%
o Yes, but there just aren't enough games written for it yet. 35%
o Yes, I do all my gaming on UNIX. 19%
o I used to think so, but now I'm not sure, since Loki closed. 3%
o I don't know, I never play computer games. 9%
o I don't know, I've never tried to play or code games on UNIX. 14%

Votes: 76
Results | Other Polls

Related Links
o Korean HancomOffice/Linux/KDE
o Loki Games
o Adonthell Project
o WorldForge
o Parsec
o Pingus
o many others
o commercial version
o demo
o Open Racer
o WineX
o Also by carbon


Display: Sort:
Linux and Games : The Good, The Bad, and The Bankrupt | 49 comments (49 topical, editorial, 0 hidden)
Games for Linux: mostly a business problem (4.25 / 4) (#1)
by MK77 on Wed Feb 06, 2002 at 07:36:04 PM EST

The technology is surely maturing to the point where it is reasonable to have high quality games under Linux. XFree86 4.0 has much of the necessary support for speedy interactive graphics which was missing from 3.3. The audio drivers seem to be reasonable. Major hardware players like NVidia even release nice drivers for Linux. Sometimes some of this is lagging a bit behind Windows, but overall, it's pretty good.

I think the bigger problem is the way the commercial gaming industry works. Shrinkwrap games for Linux haven't sold well, and any major publisher is going to take a long, cold look at the numbers before deciding to push something for Linux out to retail chains. Most likely, they're going to decide it's probably not worth the investment.

I don't want to completely dismiss amateur game development efforts, because people who don't have profit as a primary motivator will surely write Linux games, but unfortunately most amateur game development efforts don't have the slick production values or the advertising budget necessary to attract the attention of the gaming public.

For these reasons, I suspect that Linux will continue to be percieved as irrelevant for gaming for some time to come. There'll be the occasional iD software port, but few developers are as religious about cross-platform development as John Carmack.

The only way I can see this turning around in the near future would be if Linux happened to really take off in the extreme-low end computing market. Extreme low-end $300.00 PC manufacturers may want to save a few bucks on OS licensing by shipping with Linux and StarOffice instead of Windows, and if very many of those end up in homes, that could create a small commercial market for Linux gaming. But that's all hypothetical, and even it it were to happen, it would be low-end, less resource intensive gaming brought to Linux, not whizbang gore and extreme 3D explosive action associated with hardcore gaming titles.


--
Mmm... rageahol

You may be underrating the power of good code (none / 0) (#2)
by carbon on Wed Feb 06, 2002 at 07:49:52 PM EST

...unfortunately most amateur game development efforts don't have the slick production values or the advertising budget necessary to attract the attention of the gaming public.


I would disagree with you here. A good game (and I mean a really good game) is often enough advertisement enough in itself. In particular, a really good game will probably end up being bundled with distros, giving it popularity by distribution.

Also, Windows gamers who are reluctant to install Linux just to play a game often don't have to. All the really good Linux game toolkits that I'm aware of (SDL, ClanLib, Crystal Space, etc) are easily portable to Windows. It may be hard to port Windows games to Linux, but if you develop a Linux game, then it will probably comile with little modifications on Windows (and in the case of SDL, Mac and even a few consoles).

But that's all hypothetical, and even it it were to happen, it would be low-end, less resource intensive gaming brought to Linux, not whizbang gore and extreme 3D explosive action associated with hardcore gaming titles.


As any true gamer will tell you, cpu cycles used is not neccessarily equivalent to the niceness of a game. Commercial companies tend to do this so they can put out nice looking screenshots, but an OSS game develoeper, who is really on the side of the gamers more then the development team individually, will likely focus on gameplay above all else.



Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Indie games are fighting an uphill battle (none / 0) (#4)
by MK77 on Wed Feb 06, 2002 at 11:38:15 PM EST

I would disagree with you here. A good game (and I mean a really good game) is often enough advertisement enough in itself. In particular, a really good game will probably end up being bundled with distros, giving it popularity by distribution.

I don't know. Maybe a good indie game would capture the attention of the subset of Linux users who are also gamers, but I don't think that you'd see a review in Computer Gaming World. Is that the measure of success? Perhaps not for the individual developer, but I think it is reasonable to measure the success of a gaming platform by the number of people who use it.

All the really good Linux game toolkits that I'm aware of (SDL, ClanLib, Crystal Space, etc) are easily portable to Windows. It may be hard to port Windows games to Linux, but if you develop a Linux game, then it will probably comile with little modifications on Windows (and in the case of SDL, Mac and even a few consoles).

This is true. SDL is nice. Relatedly, I think PyGame is really interesting for games where performance isn't a critical factor. But if you develop a game under Linux and port it to Windows, where it becomes successful, you have done well with your game, but Linux is no closer to being a successful gaming platform. And that's what I thought we were talking about -- the viability of Linux as a gaming platform.

As any true gamer will tell you, cpu cycles used is not neccessarily equivalent to the niceness of a game. Commercial companies tend to do this so they can put out nice looking screenshots, but an OSS game develoeper, who is really on the side of the gamers more then the development team individually, will likely focus on gameplay above all else.

Gameplay is nice, but it's easier to get attention for you game with graphics. In the realtime strategy genre, I think Kohan has really nice, deep game design, but all my friends seem to be playing Empire Earth instead. Are they tasteless philistines, or did Sierra do a better job with the marketing?


--
Mmm... rageahol
[ Parent ]

Hmm (none / 0) (#13)
by carbon on Thu Feb 07, 2002 at 04:43:31 AM EST

Well, not exactly. We seem to have different metersticks to measure success by. For me, a successful game is one in which at least 20% or more who are a fan of hte game's genre have at least heard of it and played it once. It doesn't matter how many reviews it gets, or what OS it runs on, if people play the game and enjoy it, it's successful. BTW, calling Adonthell an Indie game (not that I have anything against Indie games) is sort of like calling KDE an indie desktop environment or calling UNIX an indie operating system.

Also, Windows doesn't have to be unsuccessful for Linux to be succesful. Qutie the opposite: if a game runs on Linux and Windows, a person might be more inclined to migrate for other reasons (stability, for instance).

Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Homebrew section (4.00 / 1) (#28)
by hardburn on Thu Feb 07, 2002 at 12:35:29 PM EST

. . . don't think that you'd see a review in Computer Gaming World.

Last time I checked (which was a while ago), Computer Gaming World had a special article each month titled "Homebrew Gaming", which focused on independent game creators.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
You may be underrating the power of good art (4.00 / 1) (#15)
by Ranieri on Thu Feb 07, 2002 at 06:34:17 AM EST

The vast majority of the open game projects i have seen so far tend to have rather ugly visuals. This is undoubtedly due to the fact the open source/free software community draws (by it's very nature) more coders than artists.

When looking at the credits for commercial games though, one usually sees that over two thirds of the people involved in creating a game were assigned to non-coding tasks: level design, modelling, texture generation etc.
Few project manage to get hold of decent artists, with WorldForge being the most notable exception. Even if projects do manage to get good people to do the artwork, their numbers are usually small, and they end up being the bottleneck in the development process.

Without proper artwork most open game project are destined, unfortunately, to remain rather amateurish looking.
--
Taste cold steel, feeble cannon restraint rope!
[ Parent ]

lack of flash may be a good thing (4.00 / 1) (#22)
by darkonc on Thu Feb 07, 2002 at 10:53:04 AM EST

A good game is a good game.

Even though OS gamers may not have access to the kind of hardware it takes to do lifelike animation modeling of socker players, I get a good deal of fun out of good game play. Really flashy graphics and bad game play are more of a turnoff for me than the the other way 'round.'

IF OS game writers focus on quality code and playablity, I think it's quite likely that we'll soon end up with a lot of high-quality OS gaming. It's entirely possible that an OS game will even be the genesis of a new game genre.

Never underestimate the ability of a few dedicated people to change the world -- Indeed, it is the only thing that ever has.
    -- unknown

Killing a person is hard. Killing a dream is murder. : : : ($3.75 hosting)
[ Parent ]

Sure (4.50 / 2) (#3)
by Eloquence on Wed Feb 06, 2002 at 07:56:28 PM EST

I think few people doubt that Linux is completely capable of running any kind of game that Windows can run. SDL may not be DirectX yet, but has other advantages, such as being completely open source. It's really just a question of platform dominance -- build it, and they will come. Or rather, when they come (the users) they will be built (the games). Linux on the desktop, dual-boot installations, Microsoft OEM contracts, yadda yadda yadda.

Some other free games that stand out among the rest: Heroes is an excellent Tron-like game with different modes and trippy graphics. Freeciv is an ever-improving multiplayer clone of the addictive Civilization series (hopefully they'll implement the Civ3 ruleset as well). Nethack is something you should stay away from if you want to keep your sanity.

Unfortunately, the Linux community is still lacking a decent software review index, which makes it a bit hard to find the good free games, but there are really some nice ones out there.
--
Copyright law is bad: infoAnarchy Pleasure is good: Origins of Violence
spread the word!

wrong (3.00 / 4) (#6)
by turmeric on Thu Feb 07, 2002 at 12:44:50 AM EST

"I think few people doubt that Linux is completely capable of running any kind of game that Windows can run. SDL may not be DirectX yet, but has other advantages, such as being completely open source. It's really just a question of platform dominance -- build it, and they will come. Or rather, when they come (the users) they will be built (the games).

If microsoft thought this way it would still be a penny stock. Programmers build programs, people use the 'sublayer' shit so they can run programs. They dont give a fuck what is underneath as long as it shuts up and doesnt bother them. Your idea that 'linux is completely capable of running any game that windows can run' is nearly meaningless. Theoretically any turing-complete computing device can run any game windows can run, but there is a hell of alot more to the gaming industry that @#$@#$@#$ 'running the game'. And saying SDL is 'no directx' is a bit of an understatement. Like, it doesnt even tell you when its in the wrong bpp mode and is doing super slow 'simulation' of the bpp mode. That is just blatantly unacceptable and MS would fire whoever did that if they were working there. What exactly do you mean by 'capable'? I figure you have no idea what the hell you mean, other than 'linux is just as fast as windows', which is a stupid statement for a programmer to make seeing as how almost all games are CPU-bound, and the OS has just about 0.0000% influence on how fast the CPU is running, and maybe 0.00000% influence on how fast data moves along the videobus. Sorry, linux never had anything to 'prove' in that category...................... Oh wait, you still have to do something about those goddamn cron jobs dont you?... muhahaha. Oh wait, just because it does cron jobs randomly and fucks up your game, doesnt mean 'its not capable', like i said, any turing machine can play any windows game.

if you make it easy and cheap to write, install, and run games then the games will come. This is the lesson of every goddamn console box out there. Before nintendo games hit the market, there was, guess what, no such thing as nintendo. Do you think the nintendo people were sitting around going 'we need to get more people to have nintendo boxes, then more nintendo games will come out.. therefore how do we get people to use nintendo boxes?' HELL NO. THey think 'We can sell these games, ok so people need a nintendo box, big deal, we will sell that around cost or less.'

IE PEOPLE WANT TO PLAY GAMES, NOT DICK AROUND WITH THEIR FUCKING OS AND HARDWARE FOR 30 HOURS GETTING SHIT TO WORK RIGHT.

If you saturated the market at this instant so 75% of people were using linux at home, in its current crashy buggy state with no standardized anything, then games would not sell well at all. Why? Because nobody could get the goddamn things to work right. They would bring them back to the store saying 'this doesnt work' ,the game market would crash, nobody could afford to make games anymore. Congratulations, linux zealot, you just destroyed a multi-billion dollar industry with your moronic arrogance.

This would be much like the CDROM debacle of the late80s/early90s where there was a period of time where CDROMs were getting returned to stores in massive numbers because people could not get the damn things to work because they were mostly incompatible with each other and with the computer systems and operating systems of the day.

Why the hell do you think MS invented directx in the first goddamn place? where the hell do you think they invest their 50 billion dollars? They are standardizing the goddamn platform so that there will be a game market, so that people will buy games, so they will buy computers, so MS can sell more shit.



[ Parent ]

No (4.50 / 2) (#8)
by Eloquence on Thu Feb 07, 2002 at 01:53:17 AM EST

You make two points, both of which are to some degree correct:

1. SDL doesn't provide the same kind of functionality as DirectX.

That's true and pretty much what I said. I don't think SDL is completely useless -- it's a reasonable abstraction layer without which many games on the Linux platform would not exist. With its help, several 3D and 2D games from Windows have been successfully ported to Linux, and they work, for the most part, quite well. Unlike DirectX, SDL is a public good and can be enhanced and expanded by anyone. Proper exception handling and stuff like that is typically something missing from hacks, and SDL certainly is a hack. A single game company could quickly fix such problems, though, and all subsequent game companies would no longer have to deal with them.

2. Installation is messy

This is one of the few remaining major criticisms of Linux: Installation of the OS itself and individual applications on top of it is terribly messy. In theory, both problems are solved already. There is decent autodetection, there are good graphical front-ends for virtually every element of system configuration, and there's apt-get (and front-ends to it). apt-get is a quite revolutionary concept unmatched by anything in the Windows world (there you have Windows update, which only updates Microsoft applications). Red Carpet goes into a similar direction and is probably more accessible to end users.

The problem with Linux at the moment is deployment. SuSE and Red Hat don't care about you and me as desktop users at all. The way SuSE has handled, for example, the Euro-sign addition is a sad tale of its own. It's totally understandable to get frustrated with mainstream distros because they are advertised as end-user ready, yet practically fuck you everytime you least expect it. Mandrake has a somewhat more pro-desktop attitude, but is under pressure from investors to move elsewhere. Debian is still a distro built by hackers for hackers and it will take a while until it becomes anything else. There are some promising new desktop distributions, most of them based on Debian, but these will also still take some time.

A PC manufacturer, however, could deploy Linux in such a fashion that both software installation (through either a graphical apt-get or Red Carpet) and system configuration are already well configured. Installing games on such a machine would be a simple selection and click process, even easier than on Windows (until MS integrates software sales into the OS). It may well be that one of the OEMs at some point will do this. If properly deployed, your 75% scenario would have many people surprised how simple it can be to install (and update) software.

Note: I know that apt-get isn't perfect. But there are already hooks to low-level system configuration in apt-get so that theoretically, even more complex configuration tasks are no problem. Considering that Debian is a minority distribution, the joint efforts of package maintainers (which is work which should be done by commercial distributors) are quite impressive, and it's work by motivated, often altruistic people -- exactly the kind of work that should be applauded, not dissed.

Getting Linux working properly on your own is still a somewhat sordid experience, simply because the distributors have moved away from end users at the most critical point in development. The changes that need to be made to distributions to turn Linux into a nice, easy to use desktop system are relatively minor compared to the changes Linux as a whole has undergone in the last years. But so far, nobody has cared enough to make them.

Referring again to my initial paragraph, what I meant is that Linux first has to become a viable desktop platform before it can be a viable gaming platform. Linux has no advantages as a gaming platform which would justify concentrating all efforts on making it one. It does, however, have lots of advantages as a general desktop platform, if properly deployed.

By the way, I'm posting this from NT 4, which is still one of my favorite operating systems.
--
Copyright law is bad: infoAnarchy Pleasure is good: Origins of Violence
spread the word!
[ Parent ]

YOU ARE ON CRACK!!! (5.00 / 1) (#24)
by Ender Ryan on Thu Feb 07, 2002 at 11:12:43 AM EST

You obviously have not done any serious programming in SDL, as it is not nearly as limited as you seem to think it is. While it doesn't come out and beat you over the head telling you that it is emulating a particular bit depth very slowly, you CAN tell if that is the case very easily when you set the mode.

For using for 2d video, SDL performs well enough, even without any hardware acceleration as is the case with some of my machines here.

Also realize that SDL is simply designed to make things work, across multiple platforms as well Goddammit!(put that in your DirectX pipe and smoke it!), not be the greatest graphics library ever designed. SDL ONLY does 2d graphics, and these days getting acceptable 2d graphics performance is always easy enough. Where the real work is done these days is with 3d graphics, and SDL and OpenGL can be used together extremely easily.

FYI, I play all my games in Linux, and they never crash. You seem to be implying that Linux gaming is buggy and crashy, which is simply not the case.

Sure, I may be an experienced professional who knows Linux well and it DOES have a long way to go before it is easy enough for the average Joe to run it. But that doesn't mean that it's not getting there, and it certainly deserves much more credit than you are giving it.


-
Exposing vast conspiracies! Experts at everything even outside our expertise! Liberators of the world from the oppression of the evil USian Empire!

We are Kuro5hin!


[ Parent ]

Pointing blame (5.00 / 1) (#27)
by hardburn on Thu Feb 07, 2002 at 12:29:44 PM EST

Like, it doesnt even tell you when its in the wrong bpp mode and is doing super slow 'simulation' of the bpp mode. That is just blatantly unacceptable and MS would fire whoever did that if they were working there.

Doing bpp stuff is the job of the underlieing windowing system. Which means it is an X problem, not SDL. As I've said in other posts, get rid of X and the problem goes away.

Incidently, the X standard is done by the Open Group. What company, among others, is part of that standard group? MICROSOFT!

. . . seeing as how almost all games are CPU-bound, and the OS has just about 0.0000% influence on how fast the CPU is running, and maybe 0.00000% influence on how fast data moves along the videobus.

While it is true that the OS can't do things faster than the hardware can handle, that doesn't mean that the OS can't do things to make the use of that hardware more efficent. Virtual memory is a prime example (granted, the current VM in 2.4 kernels suck).

PEOPLE WANT TO PLAY GAMES, NOT DICK AROUND WITH THEIR FUCKING OS AND HARDWARE FOR 30 HOURS GETTING SHIT TO WORK RIGHT.

Gamers will. They do it on Windows already.

They are standardizing the goddamn platform so that there will be a game market, so that people will buy games, so they will buy computers, so MS can sell more shit.

Standard for the current version, yes. However, will the game you write today work on the next version of DirectX? The past has shown that the answer is very often no.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
wrong wrong wrong (0) (3.40 / 10) (#5)
by turmeric on Thu Feb 07, 2002 at 12:21:31 AM EST

"Linux could become a commercial gaming platform as or more powerful then the ubiqtuous Windows, if only more people used it. " Why should a gamer 'use linux' if it doesnt do the damn job? If linux were really a 'great gaming platform' then the game player would not have a clue he/she was using linux, he/she would just be playing the game. Much like people have no idea and dont care what a movie like Bugs Life or Lord of the Rings was rendered with.

Sad,but true, just like you have no idea where the copper came from that is used inside your pentium and athlon, gamers do not give a rats ass what OS is underneath the hood. They care about if their character can get player-killed or if they can have booster-rocket-race-cars or if there are lots of neat fun things to do.

Sadly, linux has about as much chance as becoming 'transparent' to the user as a dump truck has of fitting in your sandbox.

I stated in my previous article on linux games, linux does not have the ability to even set a given video mode at a given bit depth without having to hand-edit the /etc/X11/XF86config files as root and rebooting X, risking a lockup and/or having to reboot. I hear the 'new versions of X address this problem' . .. . meaning that it probably doesnt work right in any of the current distributions and may never work since the maintainer might decide 'this is not fun anymore, everyone keeps bitching at me' and leave like has happened on countless other open source projects, or it might work in future releases of X but no distribution accepts it because it needs to be run as root, or etc etc etc etc etc, ad infinitum.

This is bad enough, but the real problem is that almost all of the unix bigots who run the Xfree, kernel, etc and other relevant projects do not understand the fundamental idea that you need to make it easy to do things like switch video mode. They will start blabbering how 'that is a waste of time: what we need is display-pdf and latex and display-postscript', they dont know how modern games are programmed or used any more than they know how to make ice cream.

If you want games to 'run on linux', then you have to GIVE GAMERS WHAT THEY WANT. AND THEY WANT SOMETHING THAT IS EASY TO USE AND WORKS RIGHT EVERY TIME.

better means

  • 1. reliable way to set a given video mode with a given bpp, that almost always works, or gives an extremely specific error message if it does not. i know there are alot of people who think 'why should it matter' but these people need to like, actually work on a !@#!(@#)!@# game for once in their lives instead of cluttering up the internet with their idiotic ideology.
  • 2. easy and stable video and sound API that doesnt change every 5 minutes. I know some people dont think it matters to have a stable API and 'backwards compatibility' is to them a waste of time. Of course, these people are assholes who dont give a second thought to the people who are trying to program to these APIs and treat them with no respect whatsoever.
  • 3. a single @#$@#$ binary compatible linux base so you dont have to have a 'redhat' version of the game, a 'debian' version of the game, a 'suse' version, a 'mandrake' version, etc etc etc. Can you imagine what a pain in the ass this would be for a gamer? no, of course the unix people dont they are too busy playing nethack on a vt100 they pulled out of a dumpster. GUESS WHAT, VT100 IS A HORRIBLE BUT STABLE API. Oh yeah, btw, 'statically linked' games are STUPID. First off, they dont always work! Ive used statically linked stuff that, guess what, breaks. Maybe this has something to do with mismatch of kernel with the libs that are statically compiled? No, who knows, who cares, it doesnt work and it is supposed to and I am not Joe Q LibC hacker, and even if i reported the 'bug' the libc people would say 'thats not a bug thats a feature'. Oh yeah, one more thing, STATICALLY LINKED STUFF DEFEATS THE WHOLE POINT OF THE LAST 15 YEARS OF COMPUTER SCIENCE RESEARCH ON SHARED MEMORY, PAGED MEMORY, SHARED LIBRARIES, DYNAMIC LINKING, ETC, THUS MAKING YOU BUY THREE TIMES AS MUCH RAM AS YOU REALLY NEED.
  • 4. No, requiring people to compile a game is NOT ACCEPTABLE. it is STUPID and WRONG. IT TAKES TOO GODDAMN LONG. DIFFERENT GCC VERSIONS ARE NOT EVEN COMPATIBLE WITH EACH OTHER. LET ALONE DIFFERENT HEADER FILES FOR DIFFERENT LIBRARIES.
  • 5. actual working joystick calibration infrastructure that is easy to get to and easy to set up and does not require any more than 2 seconds to comprehend and run. in windows, this means 'put it in the control panel'. in linux this means, ,, well god knows what. NO RECOMPILING THE GODDAMN KERNEL.
  • NO RECOMPILING THE GODDAMN KERNEL
  • NO RECOMPILING THE GODDAMN KERNEL
  • NO RECOMPILING THE GODDAMN KERNEL
I know alot of people think 'you need to have linux installed first, then people will play games on it' . . . This is total bullshit. People get the games they want and then worry about which platform it is on. Ps/2 is not sold 'just to have a ps/2' it is sold to have grand theft auto 3, and other similar games. Please go over video game history and notice this pattern: people buy games, they dont give a fuck what machine it is on unless they are some weirdo twisted minority who gets off on that sort of thing.

Now you want game publishers to switch from the moderately stable and backwards compatible (things that use directx3 still run on directx8.1) video/sound subsystem, a working install/uninstall system, etc, into linux, where you dont even have a standard way to switch video modes properly??? The job of the OS is to MAKE SHIT WORK, not to be some high holy hand grenade against capitalism, oh, and buy the way take 10 hours to install a video card by editing 5 different textfiles and recompiling a kernel. THE OS IS SUPPOSED TO JUST BE THERE AND DO ITS JOB, AND LINUX CLEARLY IS INCAPABLE OF DOING THIS AT ITS CURRENT STATE FOR THE PURPOSES OF VIDEO AND SOUND DUE TO A EGREGIOUS LACK OF STANDARDIZATION AND ARROGANT AND IGNORANT ATTITUDE AMONG THE KEY COMPONENT DEVELOPERS.

You can ship linux alongside the game, have a 'miniature linux' startup when the game starts, and run off of that. The only problem you have is that linux would be using a giant file on a DOS filesystem. Oh, and linux cant even come close to read/write to the NT/2000/XP filesystem. Congrats, you just made it so that every one who wants to 'try out linux' has to repartition their hard disk, which basically means risking losing all their data. boy, thats just really really really stupid.

Try learning your topic (4.20 / 5) (#11)
by carbon on Thu Feb 07, 2002 at 04:37:39 AM EST

You repeately demonstrate in your comment that you really can't be bothered to actually see how things work when they're set up properly. Comments like yours really bug me because they complain endlessly without actually bothering to get up and do something about it (aka armchair coder).

Here's a blow-by-blow, paraphrased response to every one of your complaints, with all the repetition and ANNOYING CAPS removed:

No-one cares what OS they run


Good point. Wait, wait, no it isn't. Every serious gamer I've met knows something about computer technlogoy. Even on Windows, you must tweak your system to get decent performance out of most games. At the very least, every gamer is going to know and care about what kind of video card and proc they have, to say nothing of the OS. And besides, it doesn't matter. Linux can be transparent. I can make my Linux look and feel just like Windows (though I'd rather not, since I prefer KDE). If you're going to make broad complaints like "it's not transparent", supporting evidence usually helps

X Windows can't be reliably set to a given bpp and a given resolution


I've been able to easily set X to whatever depth and resoultion I wanted easily , since RedHat 5.2! Editing config files isn't that hard, and with xf86config, you don't even have to do that.

Stable, easy to develop A/V API that is concrete


SDL or ClanLib or Crystal Space. Pick one, it doesn't matter, because X handles all the low level video stuff, and audio is handled very well by OSS/ALSA. Plus, coding in any of these toolkits makes it very likely that you can port your game to Windows.

Not needing different packages of the game for every distro


You don't. Simple as that. Most games just plain work on any distro they can compile under, and distro packagers don't seem to have the internal prohibiton against actually compiling anything that you do. Moreover, if there are distro specific packages, that generally means the game comes with your distro anyways, so what are you complaining about?

Shouldn't require user to compile game


Gee, that's what binary packages are for, but wait, you don't like binary packages. What do you want then?

Joystick calibration


Joycal.

No recompiling the kernel


Why in the world would you need to recompile the kernel? The only exception to this rule is for NVidia cards, and that's just cause NVidia doesn't seem to care too much about it's users. But Matrox and ATI cards, as well as joysticks, audio cards, and various other gaming hardware, all work out of hte box on decent distros.

Needs backwards compatability


You do have backwards compatability, when you install older versions of the libs. It isn't that hard to do, ld can run multiple versions of the same lib concurrently with no problems.

Too hard to configure hardware


Let's review the installation process on my own PC, then, shall we? Let's see, audio card isn't working. Oh, i know, why not just go into modconf, move the cursor down to es1937 (which matches the label on my audio card), press enter, and oh, goodie, it's working. USB gamepad? Modconf. Usbcore. HID. Oh, goodie, it's working. NVidia card? Oh, this requires that I compile it. Let's see, I type make, make install, move into another dir, type make, make install, restart my box, and oh, goodie, it's working. Hmm, that about covers it.I don't know what kind of system you've been using, but mine certainly seems to, and I quote, MAKE SHIT WORK.

Developers are arrogant


Even if this was true (and it's not, most every developer I've met has been quite friendly), why do you care? You're using the software, not the developers personalities. And by the way, you don't even have to pay for it! Personally, I think it's pretty dang arrogant of Microsoft to expect you to pay for software when a free equivalent is availabe, on the sole basis that Microsoft has paid off all the game and hardware developers to make Windows games and drivers.

You need to repartition the drive


Uh, it's a seperate Operating System. That's kind of implied. And even then, it isn't really, you can use various Dos-based Linuces. Plus, to say repartition is highly dangerous is really kind of ignoring the fact that just about every bit of paritioning software ever made except DOS FDISK (and this includes Parition Magic, Presizer, and even FIPS) has a decent UI that actually allows users, even idiot users, to repartition their drives without causing massive data loss.



Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
X11 (4.00 / 2) (#17)
by katie on Thu Feb 07, 2002 at 07:36:10 AM EST

"X Windows can't be reliably set to a given bpp and a given resolution"

I think he means "programmatically".

I can't write a lump of code that will go "I'd like 800x600x16 please" and have it work. Even if the /card/ will do it, the XConfig might not - especially, if like me, you set it up to do just one mode because the autoprober guessed wrong for the card, the monitor's manual omits information like the sync ranges and you plan to spend some time doing some work rather then watching your monitor making buzzing-whining sounds while you try and guess the settings for 30 modes.

So I use windows to find the best resolution, and then set that up as my only X resolution and then go something less dull.

Games aren't necessarily that happy about that, so I buy the Windows versions instead. You have to pay me actual real money to develop for Windows systems instead of UNIX. But playing games on Linux contains too much "arguing with system settings" and not enough relaxing for me to do it. Linux games haven't converted me yet; it's not surprising my mum still uses Windows.


[ Parent ]
Untrue (5.00 / 1) (#32)
by carbon on Thu Feb 07, 2002 at 02:28:52 PM EST

On the contrary, games can request any resolution they want. For instance, at the beginning of every ClanLib game is a line like:

ClanDisplay.init(800,600,16);

IF you support that resolution, X will happily switch for you. Moreover, if your bpp is higher or equal to what the game wants, it will have no problems displaying graphics, dithering down as neccessary for higher rates.



Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Yes, but. (none / 0) (#41)
by katie on Fri Feb 08, 2002 at 08:44:10 AM EST

This assumes one has correctly built the XF86Config to support that resolution.

If you haven't, it can't switch to it.

Or, if you've left incorrect lines in there, you're listening to a howling monitor and watching screen-roll wondering how to fix it.

**YOU** may well have your config set up properly. Congratulations. ID has sold one copy of the game. However, I haven't, because I can't invest that amount of time when there's an easier option and my sister and parents wouldn't because they wouldn't know how.

Cha-ching!

That's four copies of the windows version sold. (Yes. My parents do buy games.)

You see the point here, or are you going to miss it again?

Whenever someone stands up and says "Hey - you're not selling Linux games because linux takes too much fucking configuring", the reply is always "Well, MINE is fine". Yes, yours is fine. You are not the world. The world, in general, finds setting their VCR hard.

Quick - what brand of monitor do you have? Only two of my three monitors have a brand written on them, and only one of those is in the "pick a monitor" list on the X11 install.
Most of the world looks at the desk and asks which of the boxes is the monitor.

The problem with widespread linux usage is like the problem with my car at the moment.

It rattles doing anything over 80 round corners. [It has strange application errors a lot]

Yes. I could learn how to refit the exhaust properly.
[You could learn all the fiddly details about hardware you never knew existed and install Linux]

Or, since it doesn't actually happen /that/ often and it doesn't bother me /that/ much, I can put up with it until I have a spare day to take it the garage.
[I just restart it every so often, and it's OK then. I'll get my nephew to look at it when he's here at Easter.]


Maybe, maybe if we all hope really, really hard, everyone in the world will become a computer scientist, and then everyone can use Linux[1]. Or maybe not.


[1] And find the spare time to sort all these things out - I *AM* a computer scientist, and the lack of things "just working" is bloody annoying to me.

I've said it before, I'll say it again - Linux is annoying to a computer scientist. What do you expect the average person to think of it?


[ Parent ]
Uh... (none / 0) (#45)
by carbon on Fri Feb 08, 2002 at 02:31:58 PM EST

I know for a fact that adding a resoultion to your X config is spectacularly easy. You can do it any of the following ways:

  • You can just edit the XF86Config and add whatever resoultion you want (in human speak no less, no modelines) by just typing it in under the bpp you want it to be supported. Costs nothing, takes about, oh, 30 seconds
  • And if you can't even figure out how to use a text editor, you can run any number of nice X configuration wizards to do it for you. You mentioned that your monitor isn't on the list, but you ignore the fact that you can just pick the default horizontal and vertical resoultions, and there's a 99% it will work fine. You can tweak it up for more speed, but all you have to do to get your monitor, or any monitor made int he last 5 years, to work is to jsut select the default option


You say that the world finds configuring Linux hard. Guess what? I didn't. You say that isn't good enough. Guess what? It is, because I did it having zero previous experience in UNIX or with X, and it worked on the first try. Moreover, it's worked (not optimally, but it worked) on the first try for every hardware configuration I have tried, and the same is true for everyone else I know who's tried to use Linux/X anytime in the past few years. Just because you and your immediate family personally find it difficult doesn't mean that the rest of the world does.

Application errors? I never get these after I'm finished setting up my box. That doesn't mean that no-one else gets them, but I find it pretty dang likely that that's what it means, since I've installed some 5 different distros on 5 different machines.

Look, Linux isn't perfect, but for the average person (and that's who you seem so concerned about and yet so distant from) it isn't that hard to configure. It might be your choice of distriubtion (you should try Mandrake 8.2 some time, I found the installation and set up even easier then Windows) but I am sure that your experiences are not that common.



Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Still missing the point... (4.50 / 2) (#19)
by ramses0 on Thu Feb 07, 2002 at 10:11:33 AM EST

SDL or ClanLib or Crystal Space. Pick one, it doesn't matter, because X handles all the low level video stuff, and audio is handled very well by OSS/ALSA. Plus, coding in any of these toolkits makes it very likely that you can port your game to Windows.

I don't know much about programming games, but I do know that game developers like to push systems to the limits, meaning that millisecond delays from converting format A into format B add up. If you don't believe me, go buy Jagged Alliance 2 for Linux and Windows. Trust me, it's worth whatever you'll pay for them combined, they're great games. But you'll definitely be able to feel a difference (it's due to wine/2d recompilation chatter, from what I can tell. Plus sound is ever-so-slightly stuttery and delayed sometimes).

Joycal.

What button can I click to make that work? Oh wait, I ran joycal, but I forgot to compile joystick support into my kernel. And it's joycal, not Joycal ;^)

USB gamepad? Modconf. Usbcore. HID. Oh, goodie, it's working.

What button can I click to make that work? How exactly can I set it up so that the gamepad gets initialized at runlevel 3? What exactly is runlevel 3, and can you find it in windows? :^) Linux (as much as I love it) still has a long way to go in ease-of-use and/or ease-of-idiot-administration.

Surprisingly, the original author raises a lot of precise and interesting points where linux can improve. Hopefully one of the many in-development "control panel" projects will gel and gain traction, because honestly, sometimes pretty widgets *are* a nice thing. There's been talk of common XML config-parsers, and wiring up/converting as much stuff as possible to those. It's something that I'd be in favor of- and would be willing to volunteer some time for.

Just because somebody is shouting, doesn't mean they don't have a reason to shout. ;^)

--Robert
[ rate all comments , for great ju
[
Parent ]

Hmm (5.00 / 1) (#33)
by carbon on Thu Feb 07, 2002 at 03:17:46 PM EST

...limits, meaning that millisecond delays from converting format A into format B add up


You're going to have to use a middle end. This is the function that DirectX performs. Just because you have multiple middle ends that interface the same hardware doesn't mean it's slower then just having one, since they don't daisy chain or anything like that.

Oops, I forgot to compile joystick support into my kernel.


Name one kernel compiled to come with a distribution made in the last 3 years that didn't have joystick support either compiled in or modualrized? :-)

How do I make it work in runlevel 3


Umm, I hope you're being tongue-in-chic. If not, you should be aware that kernel modules have nothing to do with runlevels. Here, let me explain the exact process (on Debian, at least) for getting my USB gamepad working.

First, login as root, open up the modconf program. Move the cursor down to USB, and over to HID (which stands for Human Interface Device). Then, press enter, and Debian will install your gamepad drivers, and remember to reload them upon reboot. The user could only be confused in two areas: where to find the modconf program, and what module to load. Generally, this is about the same level of complexity as in Windows, as in Windows you must find the correct control panel (btw, it isn't Joysticks, it's Add/Remove hardware) and the correct joystick driver. Simple as that.

You're right in that graphical control panel objects are a good thiing, but what's important to note is that oftne, a good console itnerface is even easier to use. Modconf is an example of this, and another is dselect.



Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
it's not a bug, it's a feature (4.00 / 3) (#26)
by Pink Daisy on Thu Feb 07, 2002 at 12:07:46 PM EST

I'm with turmeric. I use Linux all day, but I don't play games. It's quicker and easier to reboot my computer to do that, so mostly I don't bother. Want to know what's a real turn-off though? When you make a list of reasonable and valid complaints, and get told that you just don't know what you're talking about.

Not having a tremendous interest in games, I've never got myself that frustrated, but I've done many of the same things that turmeric indicates. You should really be ashamed of yourself; what you've got is the classic case of Linux zealotry that makes some knowledgeable people write it off as a toy for unemployed hackers.

[ Parent ]
Reasonable? (4.00 / 1) (#31)
by carbon on Thu Feb 07, 2002 at 02:25:34 PM EST

I've seen reasonable complaints. Typing in all caps, repeating yourself over and over, and spending 3 paragraphs bitching over what could be expressed in one sentence is not reasonable. Some of his complaints are valid, but only after I paraphrased them.



Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Correction... (none / 0) (#47)
by madenosine on Fri Feb 08, 2002 at 06:35:09 PM EST

NVidia does not force one to recompile the kernel...I use it as a kernel module (and yes, the drivers are great,) and an XFree module

They are also availible in RPM format..

[ Parent ]
Tweaking (3.00 / 2) (#20)
by hardburn on Thu Feb 07, 2002 at 10:27:28 AM EST

Every gamer I know is willing to spend hours tweaking their computer to get an extra 2 fps or to shave 10 ms off their ping times. Considering that, I really don't see why they wouldn't move to GNU/Linux if the system offered better performance.

I don't think we'll see better performance in GNU/Linux games until we're rid of X (see my other comment above in this story).

Further, your notes about compatibility between distros is mostly moot. There is really only one main differance that affects binary compatibility: LibC. Most games I've seen have a small bit of executable code in comparison to actual game data (such as WAD files). The game data will work anywhere as long as the code is there to parse it (so you can, for instance, download a few megabytes of executable code for Quake III on GNU/Linux and use the giant WAD file from your Windows version). Recompiling the code is really no big deal.

Further still, how is the incompatibilites of different LibC versions any different than incompatibilites of different DirectX versions? (some of my older games are unplayable on a system with a recent DirectX)


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
Quite on the contrary... (1.50 / 2) (#34)
by darthaya on Thu Feb 07, 2002 at 07:00:18 PM EST

Every gamer I know would spend $$$ on buying better video card than trying to RECOMPILE THE GODDAMN KERNEL and break everything.

[ Parent ]
Quite on the contrary... (1.00 / 2) (#35)
by darthaya on Thu Feb 07, 2002 at 07:00:40 PM EST

Every gamer I know would spend $$$ on buying better video card than trying to RECOMPILE THE GODDAMN KERNEL and break everything.

[ Parent ]
Overclocking doesn't risk breaking everything??? (none / 0) (#36)
by hardburn on Thu Feb 07, 2002 at 11:26:48 PM EST

A lot of gamers are willing to do weird things to their motherboards to pump a few extra MHz out of it. Compared to that, kernel recompiles are quite safe (and probably equaly difficult).


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
Maybe you should educate yourself. (none / 0) (#39)
by Inoshiro on Fri Feb 08, 2002 at 06:47:33 AM EST

X11 doesn't impede performance anymore than the Win32 API does. DRI lets you work directly with the video card efficiently. Windows was crappy for stuff too, with people crying "we'll never game outside of DOS" -- until DirectX was released and matured. Similar problem, similar solution: let people have as few layers of thunk between the hardware and the games as possible, and make a consistent API for gaming (OS/2 had its DART and DIVE for audio and video, like Linux has OpenGL/DRI and SDL, like Windows has Direct3D, Directdraw, DirectX).

The X issue strawman is just that -- a strawman. Anyone running X more recent than 4.0.0 has nothing to worry about.



--
[ イノシロ ]
[ Parent ]
You're what we call an idiot (4.00 / 1) (#49)
by redwolfb14 on Sat Feb 09, 2002 at 02:23:08 AM EST

First off.. The only reason I'm wasting time responding to you is simply because I don't want anyone getting scared away from Linux because of your absolute lack of knowledge.

1. VT100 is not an api
2. There are many userpsace programs that allow you to change X's video mode.
3. Name a binary that was compiled on a Linux system that you can't run on another redhat, mandrake or whatever box.
4. If you don't want to compile the game wait until someone compiles it, packages it and throws it on a website or distributes it for you. IE: Loki, most free games come precompiled and packages already
5. Joystick calibration, many userspace programs do this.
6. Don't wanna compile the kernel?? Who cares.. thats what modules were made for.. Infact Redhat/Mandrake/whatever usually autoload most of the modules you need and if there is something missing.. you just insmod.
7. Backwards compatibility.. always been around, linux can't read the NT/2000/XP filesystem is a lie.. It can read but can't write.. how about telling microsoft to open some fucking specs so we can then make it work. You can't expect us to reverse engineer a filesystem without specs.. you should be fucking glad you can even read a NTFS filesystem at this point.

Basically your whole rant is a bunch of bullshit because you can't work a goddamn computer. Or take the initiative to learn something new. That's fine but don't lie and make it seem as if somehow it's Linux's fault you can't do this easily or that easily and it can't read NT/2000/XP filesystem shit.. That whole comment is just bullshit. Why don't you write a userspace program much like everyone else has for joystick calibration, why don't you reverse engineer a filesystem with no specs.. Come on man.. you're bullshit.


Say what you want because I already have.
[ Parent ]
Of course (none / 0) (#7)
by orestes on Thu Feb 07, 2002 at 01:38:56 AM EST

if you're willing to sell your soul, there's always mame (xmame).

[ You Sad Bastard ]
yeah (1.75 / 4) (#9)
by mithrandir on Thu Feb 07, 2002 at 02:55:28 AM EST

I play games on linux and I drive my microwave to work...


wojpob at wikipedia - contribute your knowledge
Eh ? (2.00 / 1) (#23)
by salsaman on Thu Feb 07, 2002 at 11:12:15 AM EST

I am sorry, I don't understand the analogy. Please could you explain.

[ Parent ]
Careful (1.00 / 1) (#37)
by Kalani on Fri Feb 08, 2002 at 03:52:14 AM EST

It's better not to meddle in the affairs of wizards.

-----
"I [think] that ultimately physics will not require a mathematical statement; in the end the machinery will be revealed and the laws will turn out to be simple, like the checker board."
--Richard Feynman
[ Parent ]
And.. (1.00 / 1) (#38)
by Inoshiro on Fri Feb 08, 2002 at 06:44:03 AM EST

Messing with retards leads to wasted time on behalf of salsaman.

Anyone who says that using Linux for a job is wrong on principle is cracked. Linux is a nice bas on which to work from; be it up, down, or just to host applications like Apache or Railroad Tycoon 2. I own and enjoy 5 Loki ports. There's no reason to not play games on Linux, just like there's no reason to not use a microwave to excite water molecules.



--
[ イノシロ ]
[ Parent ]
Ah (none / 0) (#43)
by Kalani on Fri Feb 08, 2002 at 10:25:58 AM EST

So what was the last game that you made for Linux?

-----
"I [think] that ultimately physics will not require a mathematical statement; in the end the machinery will be revealed and the laws will turn out to be simple, like the checker board."
--Richard Feynman
[ Parent ]
IANAGD (none / 0) (#48)
by Inoshiro on Fri Feb 08, 2002 at 09:33:00 PM EST

I am not a game designer; I just play many of the available games when I'm wanting a distraction.



--
[ イノシロ ]
[ Parent ]
Jumps of logic. (2.50 / 2) (#10)
by Tezcatlipoca on Thu Feb 07, 2002 at 04:34:31 AM EST

Loki closed doors because they could not make a buck selling Linux games, that proves that Linux is a viable game platform.

Argh.

-1


---
"At eighteen our convictions are hills from which we look;
at forty-five they are caves in which we hide." F. Scott Fitzgerald.
Right.... (5.00 / 2) (#12)
by carbon on Thu Feb 07, 2002 at 04:39:34 AM EST

If you read down my article, you notice the part about how:
  • Loki actually made big bucks for years before running into financial problems
  • Various Open Source games for linux are proceeding in development smoothly, particuarly Adonthell
  • Even if Loki didn't succeed, just that they ported the games shows that Linux is at least fundamentally capable of running games.


Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Tons of games.. (3.00 / 1) (#14)
by rcarver on Thu Feb 07, 2002 at 05:35:23 AM EST

Actually, there are tons and tons of games for UNIX, but most of them are crap.

Btw; check out BZ-Flag if you haven't already. It's probably one of the best UNIX games you'll ever find.

Couldn't agree more (none / 0) (#30)
by carbon on Thu Feb 07, 2002 at 02:22:20 PM EST

Yeah, most UNIX games are crap, but then so are most Windows games. It's just the ones everyone plays and talks about that are good!

And I couldn't agree with you more about BZFlag. Between that and XPilot, it's a wonder I ever get any work done :-)

Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
Hancom (3.00 / 1) (#16)
by PresJPolk on Thu Feb 07, 2002 at 07:00:43 AM EST

Hancom? Ugh. If you're going to talk KDE, try KOffice.

Hancom Office applications are NOT KDE applications. They don't look like KDE apps, they don't work like KDE apps, they don't interoperate with KDE apps, and they're not licensed like KDE apps. TheKompany and Hancom USA know this, too. They're just being deliberately misleading by slinging around the word KDE.

Hancom isn't open, but (none / 0) (#29)
by carbon on Thu Feb 07, 2002 at 02:20:53 PM EST

Korea has already been using HancomOffice for a long time, even before the theKompany deal. Now, they're switching from Hancom/Windows to Hancom/Linux. Seems like an OSS victory to me

Wasn't Dr. Claus the bad guy on Inspector Gadget? - dirvish
[ Parent ]
*nix is fine for gaming, X is unsuitable (4.00 / 1) (#18)
by hardburn on Thu Feb 07, 2002 at 10:05:54 AM EST

I think that *nix as a platform is perfectly fine for gaming. What we really need to do is be rid of the X Windowing System. XFree has done a good job of creating a free implementation of X, but they're working with damaged goods.

I think the orginal goals of X, to allow a networked windowing system, were fine goals. Some say that network-centric windowing system is not really needed anymore, but I disagree; in fact, it is only now coming to the point where it can be useful to people outside corperate networks (imagine one of those 2 GHz PCs running a server that is used by networked terminals spread throughout the house). However, X is such a shoddy system that I wouldn't want to use it to solve the one-computer-many-users problem anymore than I want it to solve the one-computer-one-user problem we commonaly have today.

In short, I would encourage anyone creating games for *nix to create the program so it will run on X now (since everybody has it already), but use APIs such as GTK+ so you can easily port the game to other *nix windowing systems. Actually, that is good advice for just about any program on *nix.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


agree and disagree (none / 0) (#21)
by Ender Ryan on Thu Feb 07, 2002 at 10:44:18 AM EST

I think X is fine. Sure, it's old, but with proper drivers it performs well and also supports networked graphics. Sure, it's not perfect, but for now it works well.

On the other hand, for portability and future longevity, I would NEVER write something X Windows specific. In all my programming I stick to OpenGL and SDL. If I used widgets much, I'd probably use GTK, but I don't really see that as an optimal solution due to lack of portability. wxWindows might be a good option though, it's a portable GUI toolkit that uses GTK(and I think other options) on *nix and native widgets on windows, and I think on MacOS as well.


-
Exposing vast conspiracies! Experts at everything even outside our expertise! Liberators of the world from the oppression of the evil USian Empire!

We are Kuro5hin!


[ Parent ]

gaming in Linux (5.00 / 1) (#25)
by Ender Ryan on Thu Feb 07, 2002 at 11:35:27 AM EST

Personally, I absolutely love gaming in Linux. I'm an experienced Linux user, so it's never a problem for me to get things working. Often I have games running much more easily and quickly than I ever did with any version of windows. I am being honest when I say that, I've had much less problems with compatibility than I ever had with DirectX drivers. Often after getting the latest DirectX drivers, old games would simply not work.

I play most of the major games of the last few years in Linux. Doom, Q1, Q2, Q3A, Unreal(using the UT engine), UT, RTCW, Soldier of Fortune, Half-Life(all of them, in wineX, runs 90% as fast as in Windows), ST Elite Force(in wineX, 90%...) and also a lot of smaller games, Rune(I absolutely loved this game, good hack and slash fun), Heavy Metal FAKK2, Heavy Gear, and I can't forget Postal ;-)

Hopefully soon I'll be playing Serious Sam (Icculus is porting it), and Alien vs. Predator.

Also looking forward to Doom 3. Gotta love id for supporting Linux(and Mac*) just out of goodwill.

Hopefully Linux Games Publishing will bring some good games to Linux, can't say I'm interested in anything they've got so far.

Neverwinter Nights looks interesting... but I'm not really into RPG's.

What gaming fun I can't get from Linux I get from my PS2. Not only do I hate Microsoft the company, I hate their OS, their apps, the windows desktop, their software update methods, their licensing practices, and the personalities that run the company. So, I'd rather live without a game not playable on either Linux or the PS2. While I don't like Sony the company much better, at least the products aren't sub-par technologically, and I get a kick out of the fact that they use Linux as a development platform... not to mention that I like the lineup of games on the PS2 the most. I'm sticking to my guns, I don't even have a windows partition even though I'm a huge game fan. I'm also putting my time, effort, and money where my mouth is and am working on some projects of my own, buying as many products as I can(supported Loki till the bitter end), etc.


* while id sells boxed mac versions of games, John Carmack has said that mac sales are negligable.


-
Exposing vast conspiracies! Experts at everything even outside our expertise! Liberators of the world from the oppression of the evil USian Empire!

We are Kuro5hin!


Give it a while. (none / 0) (#40)
by Inoshiro on Fri Feb 08, 2002 at 06:59:39 AM EST

Loki ports were high quality. The problems came when they didn't release stuff on time as promised in 2000 (they ended the year with only 4 ports, even though they had something like 9 compeleted). The games they did release were high quality and very enjoyable, I just wish they'd finished the Deus Ex port :(

For Linux gaming to be really good, we need something like a 4-year expiry clause on the code involved. That way, after 4 years of selling, the code will be released under a licence which allows the community to maintain the game. The company could still sell copies of the data files, but the engine itself won't be worth much after such a time (going by ID's example).

Loki is a good example of why an escrow of code is needed. If there's a bug in SoF's OpenGL, it'll never be fixed now. People who still want to play it will have to figure out how to maintain compatibilty with the old game (even library bugs). Very annoying.



--
[ イノシロ ]
Loki's IP (none / 0) (#42)
by hardburn on Fri Feb 08, 2002 at 09:38:00 AM EST

I don't know how all the contracts work and what parts of the game is "owned" (as far as software has owners under the current American system of IP) by Loki and what parts are kept by the orginal creator. Most likely, Loki doesn't have the copywrite on any of the orginal code, only the parts that were needed to make the port. Possibly, they even gave up that part in exchange for roaylties.

You really need to talk to the orginal creators about licensing the game code. There will probably be a lot of initial resistance, but their ears will perk up when they hear that John Carmack is already doing something similar.


----
while($story = K5::Story->new()) { $story->vote(-1) if($story->section() == $POLITICS); }


[ Parent ]
The way I see it (none / 0) (#44)
by JackStraw on Fri Feb 08, 2002 at 10:46:14 AM EST

the main hurdle now is that there's too many fish in the pond. There's only so many truly good, motivated games programmers, and too many mediocore (or worse) games that they're working on.

Soon, (or, eventually) there will be a game or a gaming engine that's built soundly and well from the ground up, and really beats the competition. All of the good programmers will hop on board, and it will continue to grow.

This seems to be what happened with KDE and Emacs, and it seems to me to be the only way the first good game will come out.
-The bus came by, I got on... that's when it all began.

I'm trying to do it (5.00 / 1) (#46)
by epepke on Fri Feb 08, 2002 at 05:05:31 PM EST

The project is called CAGEE. Phase I includes the support of image-based, panorama-based, and cinematic games (similar to The Space Bar, The X Files, Myst). Phase II will include a 3-D game engine for games similar to Deus Ex. I know some people think Phase I will not be cool enough, but not everything can be implemented at once, and I consider it a better idea to nail down the mechanisms for plotting, characterization, language, and knowledge first and then to incorporate geometry and physics.

I have already done some previous work in this area with two pilot projects, one corresponding to Phase I and the other to Phase II, from which I can take some code but mostly ideas. This includes a very decent 3-D software renderer and a considerable body of natural language, plotting, and inference engine work.

I have done considerable design work going back a year or so, though it was a bit interrupted by 9/11 and the aftermath. I have done some engineering prototyping work on the editor portion, simultaneously developing the UI in OS X/NextStep and vanilla command-line UNIX, chosen to be as far apart as possible to force discipline in doing it properly and making it as easy as possible to incorporate other interfaces.

The purpose of this project is not to make games for Linux per se, it is to make games for everything, including Linux.

I have not yet put this up on Source Forge or a similar site because I am having some trouble deciding upon what licences to use. It is necessary to make crystal clear that the game designer owns the game but not the engine and does not give up rights by incorporating the engine. The GPL, LGPL, and Artistic Licenses are all almost right, but one is not allowed to modify *GPL, and I don't want to fork too many licenses if it can be avoided.

If there is enough interest, I will format the Overview and Reasons documents for an article to submit here. Please reply to this if you would like to see this submitted. I'm not being secretive; I just don't see the point in going through the exercise just to have it hidden if there aren't enough people in this community who would be interested. After all, it wouldn't be about What's Wrong With U.S. Foreign Policy.


The truth may be out there, but lies are inside your head.--Terry Pratchett


[ Parent ]
Linux and Games : The Good, The Bad, and The Bankrupt | 49 comments (49 topical, 0 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!