Hi all.
Xith is relatively new and almost everyone (for example me :)) doesn’t know so much about it.
Why not making a open-source xith game?
Everyone could contribute with his code.
Please reply your opinions.
This has been tried quite a few times (not just with Xith). Generally it comes down to having a dedicated project manager… which is pretty difficult to find.
Kev
PS. Please don’t now get 100 posts with “I’ll do it!”. Project management isn’t something that most developers (including myself) are any good at.
Making an open-source game with xith is a great idea. 
I’ve also thought about this before. :
But as Kev mentioned there is this management problem. A manager is probably needed, but we could reduce the amount of what he has to do enourmously.
We would need to have a website, where we can agree to what kind of game we want to make, post drafts and stuff for the game.
If this is done these drafts should be evaluated and put together to one big draft. When this is done, we can split the workpower into the different parts that have to be done.
Do you think this is the way to go (if we go) ?
Basically - don’t.
It just won’t work out.
One or two of you need to get together and really genuinely write a game, properly. It’s far harder than it sounds. Your real, main problem is actually going to be finding a competent artist/modeler who can take care of everything to do with art and design in the game. Programmers don’t cut it, and if they do, it means they’re not doing the programming. Either way it’ll flounder and wither on the vine.
Cas
We could try to make a system based on votes then a volounteer writing it. 
Heck!! I think princec got a point with that - espacially I’m not very talented of doing models. But I think models start being important when you are nearly finished. So I think the basic start stuff you can do also with poorly made models.
[quote]We could try to make a system based on votes then a volounteer writing it.
[/quote]
For parts of it, right? (Everybody taking part, gets a part of the game to code)
[quote] A manager is probably needed, but we could reduce the amount of what he has to do enourmously.
We would need to have a website, where we can agree to what kind of game we want to make, post drafts and stuff for the game.
[/quote]
this won’t really help enough to make a difference (to a PM).
A project manager mainly needs you to just answer questions fast, get stuff done, make estimates of how long somethjing will take and stick to them, warn them everytime it looks like you might slip, etc.
They also need to have a single, cohesive, simple, unchanging view of what the game is and how it works. Mainly, PM’s don’t like change. Once a PM gets involved, they want to be working towards a target that doesn’t change.
More than anything, though, they need a LOT of time. It’s a heavily time-consuming job yet with little or no visilble output, so people tend to grossly underestimate the time involved unless they’ve actually done it.
I think it’s a very good idea to do a game with/for Xith3D.
I’m the initiator of the Gamma project ( not a very big project, not so features… ), and I’m doing in parallel a game ( William’s Adventures ). And progressively I add the features to Gamma that I need for the game : It think It’s a good scheme of developement… don’t know what you’re thinking ?
For the question of artists, it is really a big questions that we should answer seriously.
Do you know the Nekeme french association ? ( http://nekeme.net ) They did some very nice open-source games for Linux ( The famous Slune or Balazar for example ). I think they could help us if we ask them… They seems to have very good artists ( Look at the Arkhart artwork… ).
[quote] > link=board=xith3d;num=1110485782;start=0#7 date=03/11/05 at 15:54:47]
Do you know the Nekeme french association ? ( http://nekeme.net ) They did some very nice open-source games for Linux ( The famous Slune or Balazar for example ). I think they could help us if we ask them… They seems to have very good artists ( Look at the Arkhart artwork… ).
[/quote]
No I don’t, because I’m from Germany and don’t understand a word french.
So you think it’s possible?
Currently I’m programming a game with xith I really would like to make open source (it isn’t yet, because I haven’t got a website 
) It already got some kind of structure, but there is still loads of stuff to do. And the most important thing is, it is very open for new stuff (it’s not bound to any specific file-format, every object got it’s own java class, that contains the functions of the object and a Branchgroup,…)
So if you would like to use it as a starting point, I would really like to share the code. 8)
PS: I intend to make it to some kind of strategy game, where you can also control the people and don’t always have to play in God-mode (looking out of the sky)
You are right and while it’s possible to do that with projects at work, it’s almost impossible to force somebody into doing something in a project like this. It simply hasn’t any consequences if you fail to deliver. You are not losing money or your job.
To be honest (and being the defitist that i am… ), i think that this project will lead nowhere and that it will result in just another half finished, abandoned project. IMHO the main problem is: A lot of people don’t get stuff done. They get carried away by their real life or other, more interesting topics.
Time to prove me wrong, i guess…
It’s sure we should organize the project very good, if we want we arrive to something good.
If you want a good example of Open-Source game ( in english ), look at Glest ( http://www.glest.org ).
I think we could do something like that.
I would be happy to share my game’s gode source, too…
But I’m actually working on the game engine.
Hey hey hey, we shouldn’t let die this forum thread.
If we want to do something correct, the initiative shouldn’t stop right here !
Anyone for a basic game draft ?
+1
But before we’re going to make the draft we should decide what kind of game we want to make.
+1 for a strategy game (shooters are boring, cause they’re always very much the same)
A strategy game would be nice…
But maybe too ambitious for a first project…
Maybe a Bomberman-Style project is simpler ?
Would it be a good idea to do a network game ?
A game where several characters launch out bombs and others weapons, and it goal is to avoid the explosions. There would be no-claims bonus which make it possible to be invisible, invicibles or of going more quickly during a limited time. And the goal is to destroy the adversaries. And one pourait to organize oneself by teams.
I think that that would not be too difficult to realize, except the network part, maybe…
Yeah, … the idea is good! This way we can test also all things xith can do.
The networking part is going to be difficult, because I haven’t done networking either.
But anyways … we can do that part later, when we have got the basic structure.
Classes in which we should split the game:
Character:
- Contains the Geometry (e.g. getRootGroup() )
- Is able to plantBombs (e.g. plantBomb(Bomb b) )
- Is able to be hit by a bomb (e.g. playerDie() )
- You can set it’s position for networking (e.g. setPosition(Vector3f pos) )
- same for Rotation
- getting pos, rot also
Bomb:
- explodes after some time (e.g. explode() ) - should be private
- can be triggered( e.g. trigger(int sec) )
- has a position (getPosition() ) Position should be set with construction
World:
- contains characters (addCharacter() )
- contains a Map (loadMap() )
the creation of the above two points should be discussed - I think it could get very important later - contains bombs
- contains the main root of the scenegraph
- does the rendering
- sends input-events to player-character
I think this is good anough for a first draft. It sure needs improvement. To keep it short I’ve left the powerups out.
Do you think this is a good way of implementing it, or is it totally garbage?
PS: What’s DonCrudelis doing?
Oh sorry for my inactivity, I have been busy for a long while…
Nice ideas, everybody!
We can start with doing a little game, but we can make it possible to extend it afterwards.
I got to think about some more
I don’t have time to help but wanted to voice some thoughts.
Having watched numerous attempts in various groups to do this kind of thing I would suggest a few ideas. My favorite group is the “Clash of Civilizations” group. They have been working for years on a Civilization type game with an ever changing group of developers. Feature specific forums, general guidelines, low-level-ego’s etc have allows essentially fragmented developement to produce ever improving games over several years.Most of my thoughts are based on what I have observed there.
1 - Come up with a clear and some-what-detailed description of what you want and agree on it. Keep it simple and extend it over consecutive sets of improvements.
2 - Document the above. Break the thoughts into manageble peices and document those. This is essentially your contract with each other. It help to insure that peices can be combined into demos with minimal fuss. Small peices are the only way to work on a project with semi-dedicated developers. If possible setup a forum for each sub-project for specific targetted discussion with a Game overview forums for general discussion.
3 - Each peice is a small project in its own right.
o This allows people to work on the portion that interests them.
o It also allows a new developer to come into the process and contribute even if they don't understand the overall design.
o A solid design allows the parts to be developed separately and then periodically combined into demo's
o THE DEVELOPERS INVOLVED WITH SUB_PROJECT X HAVE THE MOST CONTROL OVER HOW IT DEVELOPS. Many of the net based games fall apart when people start dropping in and saying "OH soliders need bow AND crossbows, your model does not accurately describe blah blah. Here let ME show you the way." If people work on X they own it.
o Everyone is important in this approach. I think Magicosm took a huge death dealing hit when David had to pull out. Because he was so involved in everything. Portioned development reduces this risk
4 - Start small and only after a few weeks of discussion about goals and design.
5 - Simplify, In the game I work on in my spare time I had crowbars opening X and keys opening Y and screwdrivers opening Z. I eventuallly realized that Crowbars, hammers, screwdrivers are all different names for a Key. Something that opens something. I threw out hunks of item specific code and generalized.
6 - Produce periodic, incremental demos. Let those of us not helping at least dev-test for you. This keeps you interested because ya get to show your stuff and allows other to expose problems, offer advice, or get excited enough to join in. (At which time they are pointed to a small, designed, portion that they can immediately start helping with)
7 - A good UI takes time. Don’t get cuaght up in cool/pretty eye candy. Get the inards right and pretty it up later.
Just some thoughts your mileage may vary…
Without sarcasm, that is honestly the best advice I’ve seen on hobby/collab game development.
Hawkwind, copy and paste that post out some where, its really quite brilliant.
Kev
[EDIT: I’ve blogged it I thought it was so good]
Yeah, hawkwind is right.
My project is started since some years and I experienced the same sort of things… But he explain a lot better than me !
So I’ll see if I can start a specific-forum hosted on the MagicSpark.org site ( the group where I’m programming ), and I give you the url soon.
The implementation seems good, but following hawkwind’s advices, we shouldn’t begin to implement right now…