Why is open source game development failing?

[SiberianSniper]

[MalachiX]

Not every game needs to be a huge monstrosity. Creating open source alternatives for something like Diablo III that is just as good is nearly impossible. besides, those developers work hard and they deserve to be paid. I think focus should be on small and medium sized games. Games like Ltris, Lbreakout, bzflag, battle of wesnoth, supertux, and neverball. Large games should be secondary. Here are some of my thoughts regarding open source gaming:

- To much work is required in making a game. remember that people volunteer their time to do open source programming; and not many people have the amount of free time necessary to create a commercial level game in their spare time.
- To much division. Why make 8 versions of a First Person Shooter! Usually everyone pursues their own route and ultimately either a bad game is created or no game is created at all.
- From Scratch is bad. To many games are made form scratch. That is an excellent learning experience but the work load is to heavy to finish most games created from scratch. Look at Halo, it is built off the Unreal Tournament Engine. Look at The Witcher, it is built off the Neverwinter Nights Engine.
- Game Frameworks needed. Game engines are great, the ability to create polygons and so forth. The game engines give us legos. However what is really needed is game frameworks. Game frameworks give a blueprint for what needs to be constructed. The logic needed to create a first person shooter only needs to be written once. Then the framework needs to be so flexible that the player doesn't even notice they are running games utilizing the same code base.
- Better art. The most dismal thing is the lack of art. Most developers are not artists also. No matter how excellent the code is, if a user sees a bad GUI they are immediately turned away.
- Lack of support. Unfinished projects should be supported. Even just saying "good job, keep going"or just asking for a feature request is a way to show support. Even sending in a bug request is good. It lets developers know that people out there actually care about their work.
- Multiplayer. We really need a serious open source gaming community like Microsoft's The Zone or Gamespy. If all open source games were accessible through one major portal similar to those; then open source gaming would draw more support. It would also help the lines of communication as a whole.

[ukavi]

okay, so during my continuous daydreaming while reading slashdot articles, i just a small idea. and when i thought more about it, it appeared to have some implications which would be rather positive for OSS game development.
idea: wouldn't it be cool if you could walk around in a virtual world, and then walk into a game?

some initial implications:
-there would be no use in this, if it wouldn't be multiplayer ("wow, look at me, i'm walking through a world of portals with no one around me!")
-there would be one world with a bunch of portals, and those portals would go to different kinds of games. when you'd go through one of these portals, you'd enter a playfield matching the game's needs, get a few playing instructions, and then start playing that specific game.
-of course, all kinds of games should be able to be put into this game world, as long as they have a graphical player object which can be associated with a physical player.

So now let's add an extension engine to this, both for server and client side. This way, new games can be created based on the heavily bloated gaming system already in place. Once such a game is created and a server is running, you can walk into it using the client software and start playing that game.

Let's look back, and forget about the "walking around to find a game" thing. it's just some little thought I had, and is not necessary for the system I now have in mind, which, to clarify it a bit, simply means that there'll be only one client and server combination for all games which will ever need to be made in the future (hopefully).

And now gasp at the advantages of such a system:
-the interface with the keyboard, mouse and 3D engine, and possibly the object, walking and physics engines (as far as they are needed for the implemented game) is consistent. in fact, there's only one. forget about starting from scratch. heck, you should simply be able to just walk around in an "empty" game.
-one account for the virtual world means one login code for the entire gaming network. one login code for many, many games. one login process for precisely as much games, so the implementations of games won't need to care about a login and registration system, and having a database of players. they just see player X enters their game and will then temporarily keep track of that player's score (if needed).
-one client, one GUI. so game devs don't need to make their own GUI (which is boring), maybe just a skin for it if necessary.
-if the game systems would be made out of scripts (which is very doable since there's only one game engine):
--these scripts can be transferred over the network a player is on, so there's no more "installing" a new game, you simply (automatically) download its scripts, GUI skins and graphical objects.
--with the right scripting API, you can really hack together a game in a couple of hours.
-since a good, big platform for new games already exists, the pre-alpha stage of games is tremendously decreased. even new programmers can be able to make simple games.
-if this would become popular and a big community starts creating code and 3D models, these can easily be reused.

regarding scripting, I would like to recall Tribes1's scripting system (used for server mods and client scripts). this has spawned a pretty big community of sunday-morning-scripters which made small but relatively massive additions to the client and server, and this worked really well. the biggest disadvantage would be that the scripting engine wasn't particularly fast, so performance of, for example, AI, was rather low.

but unfortunately also some negative implications:
-controlling the amount of servers is gonna be hard. if this becomes too popular among coders, the types of games will no longer be trackable by one person, and they would need categorisation (which is bad when you want to revolutionize games).
-security would be particularly important. done wrong means a lot of people could get compromised.
-performance is going to be a problem
-again, lotsa work before we'd see any result. but once the result is there, and it works like i think it would, it'd be quite a success
-if the scripting engine is not perfect, the types of games you could create with this will be very limited.


So... any opinions about this? Does this have a chance of having a future? Or is there a fatal flaw in my setup? Would it indeed help OSS game dev? Or is it just a random, useless thought I had while reading slashdot articles?
_________________
English is not my native language, please notify me of any mistakes.

[pigeon768]

The problem with gaming in linux is that it's way too much work for a hobbiest project. By the time it's playable, it's old and for the first six months ->year, it's buggy as hell. All the people who followed it got burnt by the bugs and don't go back.

Someone needs to just make a good opensource engine (nexuiz?) and make a library out of it. And people who make art (which is honestly the hardest part of the difference between a good looking games and bad looking games) need to make their art avaible in a generic, easily stealable format. (by 'art' I mean models, textures, and to a lesser extent level design etc)

This whole non-sense people do about making a game from scratch is somewhat ridiculous. Here's a hint: it's too much for you to handle.

[assente]

The problem with games is that they aren't only software.. behind there is the graphic, the story, the game engine..
The unique OSS game that is in some way OSS and original(*) is what Blender fondation is doing.

(*)not quake3 based, not FPS
_________________
Blog
E8400, 4850, P5q

[Yamakuzure]

First of all:

[Dr.Willy]

[kronon]

Well somethings I noticed is the lack off real open/public planning and the lack of opensourced 3d/2d graphics.
These two issues make it almost impossible to write any opensource game.

Maybe a sollution would be to get sponsors for the game and then use that money to buy 3d designs, textures, music, game sounds, server, bandwith.

Some side thoughts though. Current gamers want to have a massive multiplayer experience. For this you need atleast a server and lots of bandwith. This costs money.
So some one should make money one way or another, if only to keep servers running and bills being payed.
Ofcourse there are the donated hosting solution. But if you want to have a good game, you need to have a good server and a good ISP.

So the real problem is that people want a free game, not an opensource one. And creating games always costs money, one way or the other.

So there are x problems before you can start with an opensource game
-Need a development team that knows that you shouldn't build a game from scratch
-Good 3d engine
-Good software design
-Graphics/ Art
-Finance
-Good marketing
-A website for Players/fan support for: feedback, help, boosting interest in the game, fun, etc

On a more personal note, I think that anything that relies on crystal space is bad software. Why? Because planeshift runs faster on a 450mhz machine then a 1500mhz machine, it's slow in any case and uses (think that at the moment I should say used) the xml format for maps (xml is reaaaly slow, especialy when you have alot of date. Lets for example say euuhmm.. a map?).

A game that I do like alot is Glest. I think it's one of the best opensource games ever made^^.
_________________
Please try http://www.voidofspace.com

[Dr.Willy]

By the way ... are there any good open source browser games out there? (No flash - more like OGame ...)

[trey85stang]

its real simple.. money. Money motivates people to make good games. an open source game project isn't going to be paying anyone that works on the game anything. The ones working on the project are just trying to get a mark on their resume so they can get an internship or job at a game dev, where they will not be able to work on those types of projects under their contract.
_________________
I am not as cool as you!

[energyman76b]

IMHO the biggest problems with games are the graphics. To attract people it needs a bit of beauty from the beginning. Successfull games get attention, with attention comes artists, better graphics more success.

Look at wesnoth. At first the graphics were 'good enough' - today there are very beautifull character portraits - and the maps became better a lot too. Cute animations and nice battle animation goes straight to the 'cute' centre in the brain.

vegastrike has a history with long pauses - and some graphics were and still are sub par. But it did attract people with modeling skills. Nicer graphics attracted more people - and to be honest:
a big fight with two capship+fighter groups crashing into each other, while a ringed planet slowly spins in the background looks simply awesome. Flying alone in a system with the stars being reflected by your ships exterior - awesome. Flying too close to an exploding capship and be destroyed by the wave front. Annoying but looks awesome.
_________________
Study finds stunning lack of racial, gender, and economic diversity among middle-class white males

I identify as a dirty penismensch.

[Bill Cosby]

Well, what do you guys think about the Apricot Project (a.k.a Yo Frankie)?

[ukavi]

[Carnildo]

The reason why Open Source games aren't too common is very simple: there's generally no value in a half-complete game. In order to succeed, an open-source game either needs to be complete before release, or it needs to have value while still incomplete.

Opensource games that work are generally either
1) Games that can be written by one person in a short time (Tetris)
2) Games with strong replay value (Nethack, Battle for Wesnoth)
or
3) Games with strong multiplayer (Battle for Wesnoth, FreeCiv)

[10wattmindtrip]

I think that the interest just isn't there for the game programmers. If you look at operating systems and applications, sure there's an interest. Games just haven't gotten to that point yet, I think.
One way that I can think of to at least 'help' out the oss gaming community would be to take the source engine (or quake or whatever is good enough) and just make a mod out of it. I know of some amazing engines out there that are open source: Ogre (more of a render engine, I think), Irrlicht, Chrystal space, etc. I just haven't seen anybody capitalize on it properly, yet. I may be wrong though.

Perhaps the gaming industry is too rich right now for anyone to truly want to create a Open Source game. However, I think with this DRM crap going around, people may get tired of it and this may provoke game developers/graphic designers to sway to the OSS world. Who knows? The Source Engine could be a good stepping stone to this theory. We'll see.
_________________
What would you rather believe in: A god that never answers you, or a society that embraces you?

[tuppe666]

This really is the thread that should die. PC gaming as a whole is pretty poor compared to consoles, I suspect thats as because most of it is the Sims/Warcraft/new version of familiar game version VII pink edition with the whole crack+patch system you have to go through...and now spyware. Its all a little too poor. Console is king.

That is not to say that Linux is chock full of AAA game titles because its not it has a small selection of dated commercial games...ports etc.

The reality of open-source gaming is its only just coming into its own look of this article from 2004 http://www.osnews.com/story/8146 look at the paragraph describing "projects like this have succeeded wildly. Examples include BZFlag, FreeCiv, and FrozenBubble" Whats ironic is about Frozenbubble and FreeCiv is until both of these got a version increase this year both of them were pretty rubbish IMO. I simply couldn't play BZFlag. Whats interesting is these games are still in development and improving, but the list is a million miles away from the selection and quality today...the gap has been narrowed. look at the sreenshots of a distro that can happily to any work environment on a lazy Friday afternoon http://live.linux-gamers.net/?s=games.

I enjoy linux gaming and opensource gaming its diverse interesting has its forums and contributers and little tiffs, but overall its moving forward faster than you anything with large genre gaps, games are coming through and there are several projects to watch (vdift/supertuxkart/torcs my prediction is driving games will be the big winner of this year), and quality products are rising to the top. Thats ignoring the continued development of long term favorites, and 3D apart from that company pushing its quality proprietary drivers is getting there.

...but its not there yet, and there are not enough quality titles, and well polish, but its ready for all but those boys with their dual sli setups, although I've been enjoying OpenArena on an intel x3500 chipset

[Sprotte]

Engine is less of a problem, there are quite a few good ones to choose from.

Art is the biggest problem; this includes models, skins, textures, sound, level design, etc.

To produce the complete sound for a game like Quake, you'll ideally need a couple sample cds (the General collection comes to mind, which costs 600 dollars) and a LOT of time and patience to pick and choose those sounds. You'll probably want a full time sound guy.

Making about 20-30 levels even for something old like that, is a task that will take a small team a couple years, while not allowing them to do anything else. If they are good. (If they are good, they are likely getting paid to make recreational games). Sure, your game can use randomly generated levels or maps, like Freeciv etc, but... that will make it inherently worse than having 30 maps done by a team of good level designers. The work that goes into this (or doesn't) will be immediately obvious. Imagine Diablo 2, Starcraft, Baldur's Gate, Doom 3, HL 2, Quake 4, TR: Anniversary with randomly created maps. ^^

The textures and skins... this is something you need such a lot of talent for, that it's more than likely you could and do make a living from it. It's just miles away from the typical guy's abilities. Plus you'll need photoshop practically, which is $$$. Gimp just doesn't cut it. You'll need tools and filters intended for game design. But the main problem is being a fucking good artist. You'll ideally want two of them I guess. At least. These will be partly the same guys that do your levels.

3D models... are you aware what a reasonable, easy to use (= productivity) 3D modelling suite costs?! You can use stuff like Milkshape, but it'll drive you nuts. Most people use Maya or 3DS Max. They have reasons to use these hugely expensive programs. Oh, and you'll need a fucking good (likely professional) artist. Preferably several. Again, these will be doing some of the above mentioned things at the same time. You want people who can map, model, paint, and code, and are hugely talented artists, and don't need to eat because they'll make no money working on this game...

You'll still need a couple really good coders, even if you re-use an older engine. They have to program all your monsters, weapons, items, entities.

You'll need tools to compile your levels and export your models. The good ones will probably require Windows.

On top of all that, you need a concept and an able project lead. A well structured idealistic team that needs no food and housing, is highly disciplined and never disagrees about the concept and art direction.

And after your imagined team of hugely talented, disciplined and selfless volunteers finishes their game (after a couple years), it'll only be played by a handful of people because most others play WoW, Diablo 3 etc.

[Sprotte]

Just to update my previous post,

things like GIMP, Blender, Wine etc. have become much better, and there are a couple usable open source 3D engines.

What hasn't changed is the kind of people you need to make a complete game, the sort of discipline required, the fact that no one will be getting paid even though they have to be qualified to do what they do, and the learning curve of tools like Blender or Radiant. Not to mention the time requirement.

Also, modern games require more and more effort to produce due to the increased technical overhead to get them looking like this. To create an open source game that can hold its own even against 6 or 7 year old games (Doom3, Half-Life 2) is almost impossible already.

What can be done with relative ease are probably isometric (tile based) games, using sprites instead of models, and small indie games and casual games.

In the field of 3D games, like shooters, pure multiplayer games are comparatively easy to do, because the whole extra resources needed for a single player game (creatures, levels, sounds) can be ignored. Anything where you have to create two dozen animated creatures in Blender, a pile of level design and another pile of artwork, is going to be difficult.

Finally, I believe that a lot of knowledgeable and talented people make mods for commercial games that might not be under the GPL. It's possible that this plays a role in preventing open source projects from getting the manpower they need. Simply put, modding Half-Life is easier than starting from scratch. Open source games do not usually reach the critical mass to attract developers away from the commercial games.

And lastly, the ease of pirating commercial games is, ironically, hindering the development of open source alternatives. Not even Open Quartz or FreeDOOM have been finished yet. It's been *years*, and that technology is from the stone age. But it's easier to download doom.wad than it is to create a free substitute for it.

And really lastly, open source games don't have the power of a brand name. People want the original. If you made something that tastes and looks like Coca-Cola, and offered it for free via the internet, people would STILL buy the original Coke because there is a brand, a corporate identity, an image, and a lot of advertisement behind that that your free equivalent can't match.

In short, people who want better open source games need to look at their own game collection and what they do with it, the community they are a part of (I guess Blizzard forums see more visitors than Happypenguin forums), and ask themselves some very uncomfortable questions.

You can't expect change if you are no part of it.

[poly_poly-man]

[doubleagent]

It's been traditionally thought that because games don't have a development cycle per se that they won't work as a FOSS project. Games are intended to be finished and if they work well then a sequel is made. Penumbra Overture, however, is an excellent game which went open source once they got enough money through the Humble Indie Bundle (the highest donors being Linux users), so I think now that a donation model may start becoming attractive to game developers. If this happens, then who knows what else could happen.
_________________

[gtuxed]

I saw some nice projects based on the Spring RTS engine http://springrts.com/wiki/About.

As I noticied they were fans (and programmers) of the TA.They started this around 2000 (http://www.clan-sy.com).Yea, its being 10 years and the project didn't make revolutions but its something.

I believe that the community need projects like these since we have lots of engines, lots of frameworks and many programming languages and most game projects so far looks like 10 years in the past and unfinished.

We have exceptional programmers in the FOOS world, but most of them tend to focus on just "low level" stuff...

[Muso]

[ChrisJumper]

Open-Source games don't fail at all. There are much nice open source games out there. But the competition between the commercials is for the "mass" not enough. And the "alternative people" already knew this. If there is an open-source "engine-development", that users/community easy can create some new stuff. We will see the socialized-web-2.0-Effect. And then we do care about it. That's just the "mod" way today. But all the great Mods go commercial if there are nice enough.

I think its not bad to earn some money that way.. but if its open source it make it really hard to earn money with that game for create small things or big stuff. I knew second-life is a bad example. But if you create a game like that, which engine you could third people easy use to create some jump n run, some Quest and fighting stuff like wow.. and you give a Community the power to create and verifying (again cheaters, hackers!) contend. So it would succeed. When the game can grow like a plant and the programmers just care about its physics.

Now i wrote this.. think it is actually the state with Ogre. But when you have access to these (opensource) games it would be helpful to have a single distributor-Program like Steam, as Open-Source-Community.

Its hard for me to express my thoughts in English. If you have a Open-Source-Game like Sims and others just produce content like quests, adventures, design new moves or create cars, towns or a space stations and got really money for that/or not if they don't want, then you got a game. Open Source did just fail if you create some stuff that isn't useful in the future and it got lost, erased on a Hard Drive that no one could read or use this information. But it succeed (the evolutionary way) if you won't delete it.

If i would create an open-source game, i would start with a Webpage/Browsergame that allow the users to play a Game called "create an open source game", we need the "Lego"-Effect on an Open-Source-Engine! When you got much fun during the game development and you think you still play a game, then you got it. :)

[Jazz-KP]

I believe art is not a problem. There are a lot of just-for-fun gaming project that are based on proprietary engines as mods or, now, using freely redistributable SDK. And artists from such projects don't really care about profit, most of those mods did never make it to commercial release.

In example, most "recent" fully-featured OSS FPS engine is id Tech 3, which is more than ten years old.

[ShadowCat8]

Greetings,

@ChrisJumper and @assente and all the others wondering about the capability of using a codebase/engine to be able to easily build an OSS game:

Check out Leadwerks when you get a chance.

Is that what you were looking for?

Let us know.
_________________
________________________

"As far as the laws of mathematics refer to reality, they are not
certain, and as far as they are certain, they do not refer to reality."

-- Albert Einstein