Open-Source Gaming (Mostly) Sucks!

Thursday, April 9, 2009

I'm generally a fan of open-source software. With amazing projects like Firefox, Ubuntu, VLC, and Wordpress (that list could be much larger), you could be fooled into thinking that you could run open-source software exclusively and be a happy computer user. Except when you consider gaming.

Perhaps it's because games require a cohesive vision that community-developed software can't provide. Or perhaps it's hard to marshal the development time and talent towards a project merely designed to amuse. But whatever the cause, the annals of open-source gaming are filled with badly designed, half-finished, buggy software that is painful to look at, let alone play for hours. There are scores of open-source gaming projects (much like many other kinds of open-source software) that have been abandoned for years, the only evidence of their existence being disused sourceforge pages. But games are seemingly devoid of any of the high-quality, polished projects that stand above the rest.

Even the much-touted gems of the open-source gaming world are, in a word, crappy. Projects like Frozen Bubble are filled with poorly-rendered graphics and amateurish interface schemes that are simply embarrassing. Lincity-NG, one of the most highly-touted open-source projects in terms of polish and graphics is a decent-looking game wrapped up in an interface that could have been (or was) made in MS Paint. The gameplay may be amazing (I doubt it), but there's no way to tell, since the various game structures are wrapped up in generic-sounding categories with no clear connection or intent, and there is no tutorial or single-player in-game introduction.

I could go on, but the list of gripes grows longer the further down the list of open-source games you go. For a long time, I took to complaining about open-source gaming, proclaiming that there was not a single game worth playing that was open-source and community developed (not a commercial game later released as open-source).

Until I played Battle For Wesnoth. A 2-dimensional, turn-based strategy RPG game, Wesnoth is the epitome of everything I'd come to think was impossible to find in an open-source project. It was complete, stable, and polished. The interface was clean and professional, and well laid-out to boot. The game includes a comprehensive single-player tutorial campaign, that walks you through all of the games' functions and rules, and leads directly in to one of the dozen or so polished, lengthy single-player campaigns that are included with the game's initial download. The art is fantastic. It won't push the boundaries of any graphics technology from the last decade, but 2-D, hand-crafted sprites are a pleasure to behold, and fit perfectly with the theme of the game.

So, I'm happy to admit that I was wrong about at least one member of the world of open-source games. I hope (skeptically) that it isn't the single exception that proves the rule.

I agree with you for the most part. Linux games tend to be small-scale, casual-style games for the very reasons you mention. If you look at what it takes to produce a game that has depth, like Civilization 4, the financial and organizational resources available far outstrips effort in the open source world, but also lacks the creative vision of Sid Meyer. Open-source projects do have leads, but without a way to really get everyone on the same page, the results tend to be a bit jumbled.

It's worth remembering, too, that nearly all comprehensive open-source packages (all of them, in fact that you listed, except for VLC, and even that might be) have either current financial backing, in the case of Firefox, Ubuntu, or were at one time produced commercially and have since been given over to the community (see Firefox, again, and Blender). Projects that began as community open-source projects that are cohesive and usable are rare - even Linux was based on prior work by AT&T/Bell, and has had 15 years to get to the solid but sadly unusable (at least by the general populace) state it is in.

Which brings me to my next-to-final point: It is extraordinarily common in open-source that projects have far, far more programmers than designers, software or otherwise. I'm not saying that no programmer can design software, and I certainly wouldn't suggest that all proprietary software is pretty (in a spectrum as wide as MS Office 2007 to "Bob's Awesome Legal Solution 06", with a VB interface and access backend), but as a whole the open-source community is aware of the dearth of visual designers. Here's why: for obvious reasons, the concept and importance of open-source is generally in the domain of techies, nerds, geeks, etc. Visual designers, and I include many personal friends here, could not care less that Illustrator isn't free (or even free as in beer), they just know that it's by far the best tool for the job (I write this in shame, mourning the loss of Freehand. Adobe gets a post all its own someday soon. Edit: and here it is). So they use proprietary software, but it's not as if they chose it. They simply didn't even know that there was a choice.

There's a big push, especially in the economic strife, to educate folks about open-source, but if I'm honest, I don't see a revolution happening anytime soon. Open-source will be, at least for the moment, the domain of people who care more about the beauty and elegance of the source code to the game than the beauty and elegance of the interface and art resources of the game itself.

That being said, I'm a contributor to SuperTuxKart, an extraordinarily unfun open-source clone of Mario Kart. I don't know how to program, despite being a nerd, but I do know how to make art, and so feel valuable contributing. SuperTuxKart is not enjoyable in any sense of speed or skill, but the whole point of understanding open software is to contribute what you want, where you think it'll do the most good. Go download it and play it - the website itself makes my point much clearer.

My final point: Don't be so quick to dismiss MS Paint. Though I really doubt that anyone working on Lincity-NG used it. Not when there are open-source clones available.

That's a really good point about programmers versus designers. Looking at the list of open source games, the mix of genres skews heavily towards games that can let art design and story take a back seat (FPS, 4X strategy, racing, sim game, etc). Essentially, it seems like these are 'programmer's games,' where the technology and the engine are the central focus. Witness the almost-total lack of plot-based games. In fact, that's what made Wesnoth stand out even more: it almost stands alone as an open-source RPG game (even though it is a RPG-strategy hybrid).

I didn't talk much about the formerly-commercial-now-released-as-open-source games and engines, but there, the situation is just as bad. Look at some of the engines that have released totally as open source: the Marathon engine, the Doom engine, Quake I, Quake II and Quake III. Each of those games have included, or been modified to create deeply engrossing, innovative games, with intricate plots (Marathon, Hexen, Half-Life, Heretic, Alice, Medal of Honor, Jedi Knight, Star Trek Voyager: Elite Force, etc, etc). But look at what the open-source community has done with the same engines: a deathmatch FPS, but with aliens, a different deathmatch FPS, umm... another deathmatch FPS, but with teams, and of course, there's a slightly less sci-fi deathmatch FPS, and the list goes on.

Unfortunately, the dearth of good software design in the open-source community seems to undermine a lot of the more grandiose pronouncements about the value of free and open source software development. Every time I turn around, I hear about how open source software is going to take over the world, and commercial software development is a dinosaur industry, and everything will be great when all software is developed by happy programming communes. Well, unfortunately, as far as I can see, the open source gaming situation has, merely reinforced what I always suspected: great software takes a lot of work and a lot of money, and there's no amount of overblown hyperbole about freedom that will replace the work of well-paid software designers and programmers.

Let's not be hasty. I think there's a real market share potential for a lot of open-source projects. Some are just damn good value, even if they're not damn good. comes to mind. Despite it stagnating somewhat heavily for the past couple of major releases (New Feature: Graphs are antialiased!! - notheresstillnoexcelmacroimprovement), it's always been a rather good drop-in replacement for Office for a very very large percentage of the population, even if just for home use where sharing documents isn't as important.

Some are technical masterpieces. Blender has an unbelievable amount of features packed into 20 megabytes - one of the fastest 3d modelers I've ever used in terms of efficiency, a non-linear, node-based video editor, audio processing, even a fully-embedded game engine. Yes, the interface is really really awkward, but it runs on my Acer Aspire One perfectly well (despite the trackpad, which isn't Blender's fault) and, again, 20MB. The TRIAL version of 3ds max 2010 is 800MB. Is Blender better than max? No, of course not. But it runs natively on my Mac and my linux computer, interchanges between the two and has what I need 70% of the time. I've been a 3ds max user for over 10 years, and I still use Blender for a lot of stuff, because it takes longer for 3ds max to open than it does for me to fire up Blender, make the quick change to texture mapping or whatever, save it and close out.

And for some, they are truly superior to their commercially-backed equivalents. Asterisk, Apache, Firefox, etc. These pieces of software need the campaigning to counteract the marketing of Microsoft, say. Apache still has a higher user share than IIS, and for good reason. It's about a billion times more stable, and patches to it don't require a reboot. How stupid is that when you're trying to guarantee 99.99% uptime?

Finally, there is a surprising amount of open source software in niches where no commercial alternatives even exist, including embedded kernels for ARM and other non-x86 processors (not even Apple supports PowerPC anymore), statistical analysis, scientific data collection and processing, a ton of middleware, and perhaps most famously, Networking software such as port scanners.

I heartily nod my head in agreement as re: Nexuiz, OpenArena, etc. The best things about those games have all been cribbed from the original games they're based on. They might add lens flares or something, but it's hard to get any sense that the game itself is different from Quake 3, and often, as you say, with even less plot.

But for games that are designed with no plot in mind, like Flight Sims (some which are, apparently, even more accurate in flight physics than Flight Simulator (and still being worked on. This is one area where the commercial software disappointed fans by stopping work on the project suddenly. But don't worry, says Microsoft. You can still keep buying the last release!). Games like HedgeWars, a Worms clone with graphics that are surprisingly good (and cute) for an open game, don't need a plot, and would not benefit a lick by one.

I think there's a market for open-source games that are full of plot and sophistication (even if not in the graphics or sound), and I would remind you that there is quite a large community dedicated to that very idea, one that I have been mulling over joining and contributing to for quite a few years: Adventure Gaming. Again, one of the best games in the open universe (and in my opinion, ever) is a release of original source code: Ur-Quan Masters, but there are still a few games which have depth that certainly rival the plot of even showcases of modern commercial gaming, like Quake 4 (Plot: go somewhere and kill aliens), and Call of Brothers Warfare, 1944 (Plot: kill nazis and italians. Perfect for getting your aggressions out in a PC way if you hate germans and italians. 'No, It's 1944, so it's ok. Or whatever zombie shootkill game is released. I'm not saying these games aren't fun (I personally really like the zombie genre, one of my guilty pleasures) but I stopped playing Quake 4 about 2 hours into it and have never felt any need to continue. I've never played a WWII shooter. I'm simply saying that the commercial gaming world "cheats" in terms of plots too, but it's often better masked by a pretty face and DirectX 14.3 ShaderPixels and fluid simulation.

If open-source software was a boss for a video game, the flashing red 'weak point' would be... video games. Will there ever be a world where only non-commercial software is used? I don't think so. Things are worth what people are willing to pay, and people are willing to pay for good games. But I think of the open-source gaming situation as a bonus more than I think of it as an industry in and of itself. 'Linux also has games!', rather than 'Gaming on Linux sucks.' One of the first things I do when customizing a new Linux install is remove the games that come with it.

All of which are better than the games Windows comes with.


Rodneylives said...

You need to look into these games:

Dungeon Crawl.
Simon Tatham's Portable Puzzle Collection.
Anything by Kenta Cho.

Eldritch said...

JohnH, thanks for the suggestions. The roguelike games are classics of open-source gaming, and Tatham's puzzles look like nice versions of perennial favorites.

They are, however, pretty emblematic of one of the peculiar facets of open-source gaming: the games tend towards programatically-generated content and lack handcrafted art or plot or scenarios. Now, of course, Nethack has been well-loved by generations of gamers, but I find it striking to see the difference in presentation and design between open-source projects and their commercial counterparts.

I'm definetely going to check out Kenta Cho's stuff, though. That seems very interesting.

Unknown said...

I think your conclusions about the status of open-source gaming may hold true for a lot of projects. In general, they suffer from crappy art and tend to be clones of better-known (and better known) games, but I don't think this is a necessary limitation of open-source gaming. It has more to do with the culture of open-source in general.

The people who make these games are almost always programmers first and foremost. They aren't artists or designers, so art and interfaces are secondary. Usually these guys (and lets be honest, they usually are guys, unfortunately) are writing these games as practice for programming skills because if they had the experience already, they would be doing it for a living.

However, I think this is changing to a large degree. As games become more of an accepted artform instead of being viewed as a toy, people who aren't programmers to begin with are starting to create games. Recently developed tools like GameMaker and Unity enable the creation of games without a significant amount of programming.

We're already seeing some awesome examples of this. is a nice resource for examining the goings-on of indie gaming. The forums, especially, are a good place to start.

Eldritch said...

fergk, thanks for the comments. I hope you're right about the future of open-source gaming. I don't want to come across as too strident, because I honestly feel like the open-source, community-developed model has a lot of promise for gaming development. There's something very intoxicating about the idea of a game that continues to grow and evolve for years after its release, not just forgotten and relegated to the bargain bin like most older commercial games.

And the plethora of community-developed content (mods, maps, scenarios, etc) for many commercial games would seem to suggest that non-professionals can create a lot of great content. I'm just puzzled why more of that energy hasn't been directed towards wholly community-developed, open-source projects. Hopefully we're at the cusp of a change in that regard.

Pseudonym said...

Can I just plug Hedgewars again? I don't contribute to it (the one I DO have assets in looks pretty bad, visually), but it's got graphics and style that at least equal the graphics of many independent games.

Anonymous said...

Oooh, Hedgewars.

Unknown said...

On the other hand, as someone who has played through Wesnoth in a couple of builds now, it has plenty of problems itself, mostly within the way that the game scenarios are structured (it basically requires the save and reload method of gameplay to win on higher levels, just because the variance in damage is too unpredictable). It's not always stable, and it tends toward scenarios written for folks who are already hardcore gamers. Even on easy, it can take hours to tromp through a level. The art is OK, and the dialogue can be atrocious.

It's a good game, but not a great game. It's only so exemplary because the vast majority of open games are so much worse and because it doesn't get compared to professional games in its genre. If you had to pay for it, it'd be a B- game at best, at least for macs.

Loïc said...

You talk about the appaling graphics some OSS games have, but isn't the games engine a good culprit? We hear about "great" OSS engines for games, yet are they really appealing for artists?

Take your Frozen-Bubble example. The game resolution is so low it's a disgrace. Get the code to support SVG graphics and you'll see good artwork in no time (actually, Frozen Bubble artwork is not the problem, if the game supported SVG with good transparency and filters, the same artwork would look great).

Racing games? Have a look at Gran Turismo 5 or any modern commercial race game engine. No wonder artists won't work on OSS game contents. Would you ask Van Gogh to work for days knowing you'd only allow him to display his work in 640x480, or, if it's a 3D game, that you'd limit its models to a number of vertex 100 times lower than the average commercial game? And of course no good shaders 3.0 or 4.0 support...

Creating 2D or 3D contents take an awfull lot of time (and skills). Which good artist would spend that time knowing he'd be so limited by the game engine (or the developers restrictions on number of polygons, etc...) that his work will never look good anyway?

Unknown said...

Open source Racing Games is far - far behind commercial games, while Gran Turismo had 1000++ amounts of cars, Forza motorsport very impressive graphic, Need For Speed has high adrenaline view, but that's just a second thing. The first is gameplay, that's what make me (race driver wannabe) race and race again

I can remember gran Turismo very well, you must do license, buy car, upgrade it, and then come the biggest challenge, the endurance (2 hours plus length racing) for some people it's just waste of time/exhausting/boring/etc, but that is the the things who made me and the other GT fans race for more

need for speed is known for they cat and mouse racing, but hey! That's BAD ASS! And that's can only happened in need for speed

What about OSS games?

Super Tux Kart - looks like elementary school boys game (sorry for being rude)
Torcs - graphic is very outdated, then they only focused on ROBOT/RARS racing ( it's great if you're s.hawking)
Speed Dreams/Vdrift/Stunt Rally graphic is good and still acceptable but somehow they lack of gameplay, IMHO what they (last three) need to do is polishing they rides more, and add more friendly menu

For Speed Dreams they planning to change their anyone, and I'll be wait and how the result

Post a Comment