No announcement yet.

If you were to create a FLOSS game, how would it be?

This is a sticky topic.
  • Filter
  • Time
  • Show
Clear All
new posts

  • divan
    Hey everybody,

    I have started a project on my own so far, which is not yet open source but I will make it open source sometime. It is a RPG game with singleplayer and multiplayer parts. It is 2D but can use OpenGL for acceleration.

    Check my website at :

    This is my first LAN game and I have only been working on it for 6 weeks, mostly creating the interface, graphics system and other basic stuff. Please let me know what you think.

    Leave a comment:

  • Thetargos
    Interesting. Jedi Knight II was a fine game. Seems like you have it very well defined (in concept at least).

    Going back to the Diablo-style game, I just remembered one of my favorite games, for which I even have the original Linux CD...

    If Epic ever releases the UE 1 and the editor's source, maybe we could think of a game similar to Rune with a Diablo-style game play, mix (if you will) some elements from Wheel Of Time (another Unreal 1 engine based game) and you even have the spells added in. Change the way "regular" and "alternate" fire work to suite more in a style of Diablo II (where you can devide the spells in the buttons for attack spells and "magic" spells)... The game you've got conceived could also benefit from that, you could still use any weapon you'd liked (melee or ranged) and still use "spells". The hardest part would be to add the RPG and character development elements, though (I think of those as they are in Diablo II, flexible, yet limited, making it easier to actually build). Then there's the whole other RPG elements of hits, "dice" formulae, item's properties, item's extensions (like sockets in D2, Light Saber enhancements in JK2, etc)... Since there are already "precedence", it might actually be possible to pull that off with either engine... I think...

    Now, if only Epic were to release the Unreal 1 Engine... Anyone know why they haven't released the engine?

    Leave a comment:

  • Malikith
    You know, now that you mention UT 99, I just thought of something, what if Epic ever releases the source code to the Unreal 1 engine? That would be very cool. I think UT 99 is more flexible than Quake by far.

    Another game idea I've thought of, and probably my most wanted, and probably most thought through thus far, is a game that has gameplay like Jedi Outcast. I loved that game very much. Granted I don't think you could use lightsabers and stuff and give it that Star Wars theme, but you could use things that are similar like calling the lightsabers "Photon Blades" or something.

    A 3rd person game like that would be very neat to play, thats for sure. I'm going to say use the Quake 3 Engine again just because a game like this would definitely fit into that engine without a doubt. Jedi Outcast and Jedi Academy also used the Quake 3 engines. I'll explain how it would work.

    You could practically take say, the Open Arena game code that makes the game Quake 3ish, and modify it. Turn on 3rd person, make the FOV default at 100. Add the Photon Blade to the game, add in a substitute for force powers, call it something else obviously. Guns could later be added to the game but the main focus would be adding moves and balancing the Photon Blade.

    The gametypes, well nothing new, Deathmatch, Team Deathmatch, Duel, and CTF. So no coding needed there, its already been done. This game would probably go together quite nicely with even a very small development team.

    So lets recap this, basically what would need to be done (To hit 1.0):


    Team Deathmatch
    Capture The Flag
    Instagib Capture The Flag

    - 3rd Person
    - Photon Blade and 9 guns
    - Servers can enforce Photon Blade only
    - Admin mod built into the game for easier server management
    - 5-8 Player Models (Depends)
    - Special Power Abilities (Force Powers Basically)
    - 4 DM/TDM Maps, 4 Duel Maps, 3 CTF Maps
    - 3 Photon Blade Stances, Light, Medium, and Heavy, Light being fastest but least damage, heavy being slowest but most damage.

    - 12 Powers Abilities:
    - Neutral: Jump, Speed, Push, and Pull.
    - Good: Healing, Absorb, Protection, and Trickery
    - Evil: Grip, Lightning, Drain, and Rage

    Under the hood:
    - Based on ioquake3
    - Modified GTK Radiant packaged with game (Just to add support for the game and ease of use)

    After 1.0 you would just addon, more player models, more maps, and maybe even add in Dual Photon Blades and Double Bladed Photon Blades. Who knows.

    But thats my idea on that game.
    Last edited by Malikith; 08 November 2007, 02:50 PM.

    Leave a comment:

  • Thetargos
    I can't speak of the Quake engines as I have no experience working with them. I don't have the slightest clue how do they manage models, levels and so on. I have more experience using the Unreal Engine, but up to UT99, as I've never used the UT2K* editors (no Linux version available, and by the time UT2K3 came out, I was a looong time 100% Linux). Using Diablo as the example (as it seems to be our common ground here) and looking at how the different engines (UT/Quake) handle different stuff, I'm not so sure these engines "will cut it", mainly because they need a series of stuff before hand to even display a level, and the load times get longer and longer the more items you have in the levels (geometry, pickables, enemies, etc).

    I don't know how did Blizzard or Bethesda managed to tackle down the problem with dynamic content loading for their top 3D games such as Oblivion and WoW (I don't have experience playing WoW, either)... However I wonder if the engine (a la ET:QW) could be modified so that to have single HUGE levels (in essence, worlds) and to be dynamically loading content of the world (world geometry, placeables, pickables, loot, random objects, enemies, destroyables [like ruins or the like], etc). The trick here would be memory management and how to ensure that memory be freed under certain circumstances... The image I have in mind would be like a "window" on a 3D plane... For instance if like the world is a sheet and what you currently have in screen is a small square of that sheet "bumped up" from it causing a "deformation" (a la Matlab logo), the sheet would be the world, and the "bump" would be the memory buffer ever shifting depending on where the player moves to, freeing the equal amount of memory as it requires to load new stuff... This has the problem that it may hinder performance due to disc IO operations (as disc reads are MUCH slower than RAM reads).

    Another approach is to use what you suggest, individual level loading a la Diablo 1, and only adopt the dynamic loading for random objects which will be selected when they are loaded into memory from a library of possible objects, like monsters, loot, chests, etc...

    Of course all these "could be's" are mere speculation and it would be up to the programmers to try different models and engines to see which fits the theme and style of the game best (yet again, planning)

    Leave a comment:

  • Malikith
    I totally agree with everything you said, the work and time definitely takes its toll hehe. I mean there is alot more to think about when building a project such as a complex open source game, like a website which a nice one takes some time, the integrity of the group. I believe that even if you only have 2-3 hours a day or a little less to do some work, I still think you can still deal with the community on the side. However, that also depends on how many guys are working on the thing.

    I think the thing that will help make open source games really come to life is for that one project to come along that really just redefines how its done. We've got some really really good projects out there already that in time could do it themselves, like Battle of Wesnoth. I think thats probably the most complete open source game yet. Its even got single player which is pretty impressive. Single player seems to me like something that would be difficult to pull off in a open source model, but they made it modular (aka the campaign system).

    So I think if you were to take a FPS, for instance, it would have to be done the episode route like Duke Nukem 3d, or Quake. And have it ready for community made episodes that can easily integrate into the game. In a Diablo styled game in a Quake 3 engine, well, my only concern would be the size of the levels and load times. Because the levels will all have to be decently sized, meaning no loading every 2 seconds.

    So the server would have to basically load a ton of maps right from the start and somehow, allow the players to be in different maps at the same time on the same server. I think the best way to do the levels would probably be just do them the Diablo 1 route, each level of a dungeon is a separate level. The nice thing is, since the game would be top down, it would allow you to use less detail on the level, and be able to cut things out that are not visible, like the sky or ceiling.

    You could have a camera that spins in a 360 degree angle but then you'd have to use more detail. Which would slow things down on older systems and would require more work. So in my opinion, a locked camera top down would probably be the best route.

    As far as 3d modeling, well, whoever would be doing that would have their hands full haha. All those separate items that would attach to the body of the character, all those unique ones, all the decorations and the hardest part, the player and monster models. That would definitely be the hardest part of the development. Granted the level designer could take care of alot of the decorations.

    With level designing, I would make sure GTK Radiant is the only editor used. Since itself is open source and is in my opinion the way to level design on any Quake engine. Unfortunately, alot of people like Quark, I hate it. Open Arena level designers use Quark. The problem with Quark is that when you save a map with Quark, it breaks compatibility with GTK Radiant and you can never open the maps in Radiant after that, and Quark is Windows only. Which causes in my opinion, a huge problem.
    Last edited by Malikith; 08 November 2007, 01:20 PM.

    Leave a comment:

  • Thetargos
    started a topic If you were to create a FLOSS game, how would it be?

    If you were to create a FLOSS game, how would it be?

    Let's keep the other thread on-topic, and as such that's why I started this other thread, so we could discuss every thing we can think of about an Open Source game. I'd like to take off where Malikith left, hence a quick quote:

    Originally posted by Malikith
    Yeah art is always a issue, in my opinion though, at least with my experience with the Quake engines, 3d is easier to handle than 2d. Problem with 3d though is that you need 3d modelers, and well, good ones don't come everyday, at least in open source. I mean take a look at all the 3d open source games, I haven't seen one yet that is fairly mature that has 3d models that come decently close to a retail game nowdays (I'm talking like games from 2004 and beyond).
    Indeed... And is not like there aren't any artists willing to participate on projects (like your point bellow), it is about communication between project members and the community (so they can even get more help if needed) and fully understanding what can be expected from the project. In practice, that's VERY hard to do! The project has to be properly advertised amongst the community in the hopes to get as much attention and volunteers as possible, but before doing that, proper planning has to be in place, visual design concepts have to also be in place (so the artists know what the content should look like), and if not a schedule for milestones, a projected route with specific goals. Again, that's much easier said than done. It is easy to imagine how to do it, but very hard to actually DO it.

    Originally posted by Malikith
    Unfortunately I'm not a coder, only things I can do with the Quake engines is create levels, but I'm not bad at it. I actually made a level Open Arena but I pulled it because I didn't like the developers very much. It was mainly due to the fact I didn't like where they were bringing the game as far as art (since I hate anime with a passion, but thats my deal), and I didn't like their attitude towards the community. I mean you couldn't even find half the developers, they were in their own private little den somewhere while the main one just sat alone in IRC.
    I don't think that a "project manager" has to necessarily be a coder, it is a plus to know how to code (for code review or to modify something you spot, etc), but it is necessary to have a clear idea of where the project is going, and how is it progressing. You can't ask to meet goals which are too ambitious right from the start, not in an Open Source development model, particularly where the people are only committed to the project in their spare time, they've also got lives, is not like the project will have to ask these people (project managers, lead programmers, etc, included) to forget about their life (social, familiar, everything outside work) to work on it... would be absurd. However this also means that progress will be frustratingly SLOW. If the people working on the project know that beforehand, and if there is good communication with each other, a project like this could indeed be pulled off. Again, even if the project won't "capitalize" any monetary goods, it still requires a heavy investment: Time, that's why sponsored development like that of the Linux kernel and other "big projects" would be ideal, particularly if one or more players of the gaming industry would be willing to do it... Like what it happened with CS and Valve (of course CS was eventually "engulfed" by Valve, but the process was similar)

    Originally posted by Malikith
    So I guess my point is, would I ever do something like this in the future? Yeah, but I'd also want to be surrounded by the right people at the same time.
    I understand what you meant there... (before hand I ask for forgiveness from those ATi users that I may offend with the following comment). It is simply hard to keep the enthusiasm high when you are taking a long time to get to your goal, just like we've seen with the fglrx drivers. It is also hard to keep people involved and willing to participate. Their main priority would be their works, family and friends, and then the project.

    Originally posted by Malikith
    But back on track on everything, RPG's are harder to make than FPS's, which explains the large amount of fps open source games. I do see a few open source mmorpg's out, which is nice. But unfortunately I don't know if they will survive the test of time. Because it takes alot of maintenance and baby sitting and money to build a open source mmorpg. But I do hope they survive because I've had some fun with some of them.
    I think that much more planning has to go into an RPG game, be it an MMORPG or a single player with multiplayer game. There is quite a LOT of work that has to go into it, particularly from the planning aspects of the game. What would be easier to get done? A game similar to Diablo with a main adventure, a limited RPG aspect and focusing on action and story telling rather than RPG elements, with direct-connect multiplayer action (i.e, no need to connect to a central server to be able to play multiplayer games, and be able to do it in a LAN)? Or an MMO type of game where the complexities are greater, as there is a persistent world with any number of varying people connecting to it from different parts of the world with no central plot, but still some story behind the setting and "universe" and offering compelling adventures for the people joining in?

    One obviously requires much more infrastructure than the other, but at the same time has "simpler" gameplay elements as its goal would be to also promote social interaction among the gamers... Not to mention the whole deal of balancing the game, weapons, armor, items, damage, creatures, loot, levels, quests, plot, economy system, etc, etc. An FPS indeed is to a certain point "easier"... The new trend seem to be FPS games with some RPG elements and social interaction that raise the bar for these types of games (ET, ET:QW, CS, TF[2], Battlefield, Call of Duty, etc), effectively in some cases turning them into First Person Role Playing Game Shooter kind of games. In any case, the key element which most never be forgotten when creating a game is to keep the game FUN (which seems to be increasingly hard to achieve nowadays also). Creating a complex game like an RPG or Adventure title, with the Open Source model, for and from the community would be and is, extremely difficult.

    No wonder game companies have so many people devoted to any one title (Director, project manager, department managers, writers, artists [2D/3D], programmers, sound engineers, etc, etc), trying to "emulate" such organization is very difficult, and more so depending only on people's spare time.

    Originally posted by Malikith
    About the Quake 3 engine, I noticed with cg_shadows 2, that there is nearly no performance impact if the shadows are viewed from a distance, but up close they cause the framerate to drop like a rock.

    In my opinion, engine wise anyways, I would either use ioquake3 or the Alien Arena Quake 2 engine. Their Quake 2 engine is actually pretty good.

    Hope I made sense hehe.
    Yeah, you made perfect sense. Working with the Q3 engine would be one way to do it, although there are other engines which could be used, like Ogre, Crystal Space, Dark Places, etc. The main challenge would be to make the game (using any of these engines) scalable enough to be played on perfectly well supported with Open Source 3D hardware, with the addition of some nice touches for those with more capable hardware or using binary blobs, but I definitely think that the efforts should target platforms such as VIA's Unichrome, Intel's IGP's and R300/400 with Free drivers. Hardware balancing could also be VERY difficult to attain.

    I guess that the bottom line is that in order to get such a project going heavy planning must take place before any pixels are drawn or any lines of code or plot are written. Many hours and discussions and concepts must be put on the table before actual work and recruitment could take place. At least, that's the way I see it... With passionate enough people, good communications and clear goals, I think it very possible to pull a very complex project with the Open Source model at its core. I certainly hope some one will step up and actually have the nuts to do it.