Okay first off my name is Shane, I’m a long time web developer and short time java programmer but I’m picking up things quickly and can say I’m heading into the intermediate “level” of experience. I’ve been shooting this idea I have in my head for awhile but not sure how many lolz I will get in response(which isn’t my intent). So my idea is to get a group of people and collaborate on the game for the pure purpose of learning JME and learning from eachothers experiences and solutions.
The way this would work is use a service or online program (SVN, github, gitorious,etc)that would allow anyone to fork the code write their proposed solutions and then we can all disccuss the code that is being uploaded with real critique and build upon the code and maybe eventually build a cool open source video game for other beginners to learn from. (learning for beginners by beginners)
I don’t want to set the path for details in this post about if the game should be single player or multiplayer, the story line or what type of shaders should be used, these details should be suggested by others and then everyone else(and me too!) will expand and improve upon pluaseable ideas.
I recommend that we should stick to plauseable things that shouldn’t take longer than 2 months to implment just to keep activity and progress rolling but also 2 months is a good target to allow others to chime in when they have the free time to contribute.
So where shall we begin?
Yep, and that may be because doing building equal new node means that you are maki tong the building aa node rather then a type of mesh.to fix this you would probably need an array of nodes that inherited from the rootnode. This would mke it so you would not he. Bunch of nodes just called “node” or “building” and instead it would generate node0, node 1, etc… which would then link to the building that had the same corrosponding number.
You’re over-complicating the problem.
You never initialize the array. Think of it like this:
A normal object reference is a bit of paper that you just write the address of the object on.
An array though is a book, with each page being a bit of paper you write the address on.
When you create the object you get a “free” bit of paper with the address. Node x = new X();
But you don’t have an address book yet. Node[] nodes is null;
You need to create an address book of the correct size Node[] nodes = new Node[10]; for example would have 10 pages in it.
And then the first page exists so you can write into it node[0] = x; (or node[0] = new X();
Arrays are fixed size once created so you may be better off switching to java collections. In general you would find the collections API gives you more flexibility, more control, and more possibilities (and is a must learn for all java programmers) although there are (very) occasional reasons to use arrays instead.
Going to add NightMare’s update to mine, and also add in some fog. I’ll pull when done.
EDIT: Uploading now. But we still have a few concept problems, like why are the buildings in the middle of the road? How about, you are so high in the sky that you cannot see the sky for fog, and you can only fly a certain distance downwards. We can clip this distance with a plane. There is a plane collision shape. If we do that though, we’ll think about it after the camera has been implemented
EDIT: looks like i’m having some trouble updating some stuff
Assets updated, i like the updates, I’m already learning some better tricks. I’m having trouble merging one of the pull’s which is requiring me to use the terminal to merge it, when i get back home I’ll have to do that. I think its just because the merge is out of date after approve another branch no biggie.
I have included Nightmare’s update in the ‘Big update #1’, so unless there is something I overlooked there will be no need to update it
Perfect thanks!
Okay, a quick little question… What do I need to do, to make certain that my local reposotory from git is up to date ?..
In svn, it’s update to head.
I think one way to do it is delete your current OCLP directory, then go onto the main OCLP on the github site, then click clone in windows in the top left.
There’s probably an official way but…
I’ve never used git much (SVN for me) but there should be a “fetch latest” type command depending on what client you are using.
Well, it would be a rather fun project to try to work together with somebody on a game :)… Although I will image that it, would be a little game, perhaps a game, where it would be possible to use the physics engine in jMonkey Engine.
So yeah, you can count me in, if this game is going to be hosted on an SVN 
It’s a decent idea, those sort of peer learning projects can work. It does need someone with a lot of time and good personal and organisational skills to keep things organised and on track and get things going in the first place though.
I’m listening. Keep me updated please. @Zarch is right, we would need somebody to organize and lead the game …
Excellent, no chuckles yet 
I can understand the need for a manager to keep things on track and keep progress moving.
I originaly thought, scince this isn’t really a professional endevuer and just a learning experience we could manage each other(and our selves) and use a Project manager software to keep the projects “RoadMap” clear to everyone that joins but I think confidence might be higher with someone with a Managment title above this projects’ head.
I for one don’t think I could claim the management title but it is up for grabs.
I will set up the collaborative software on my web host when this project has produced a: Storyline, and gameplay conecept ideas.
I think those are the two things absolutly required before we can even start writing code, modeling or doing anything.
I can tell this is already going to be fun!
I think scope will be your biggest issue. If you are trying to learn coding/game-development then you should keep the idea simple and focus on the development. Lack of momentum will kill the project faster than anything else.
One approach is to write a simple asteroids clone or something and then branch out from there. It seems with these sorts of projects like there ends up being X number of months of brain storming that result in a bunch of cool undoable concepts.
Take the amount of time it will take you to really nail down your concept (completely) and then multiply that by 50 or more. That will be the length of time it takes to develop it… minimum. And that’s if you really work at it every day. In my opinion, it’s better to start simple.
If you are having trouble figuring out how to start then storylines and shaders aren’t really going to answer that question.
@nightmare2real said:
if this game is going to be hosted on an SVN :)
You might start rethinking that when you start committing art assets. Anything large and SVN will crawl. GIT really shines here, and it's not that hard to learn.
well, I gues that I will need to learn a little about GIT then as well
However for the game, I have a little proposal. Which I believe shouldn’t be to difficult to make, and yet it will introduce some of the challenges, you face when building a larger game.
The idea is, that you are a fighting cube, a cube with a sword. Which is off to fight against other cubes with swords (pvp perhaps, or against some mob AI, and along the rute it should be possible to pick up other weapons/items.
What do you guys think about the idea ?
@nightmare2real said:
well, I gues that I will need to learn a little about GIT then as well :)
However for the game, I have a little proposal. Which I believe shouldn't be to difficult to make, and yet it will introduce some of the challenges, you face when building a larger game.
The idea is, that you are a fighting cube, a cube with a sword. Which is off to fight against other cubes with swords (pvp perhaps, or against some mob AI, and along the rute it should be possible to pick up other weapons/items.
What do you guys think about the idea ?
I like it, its like the simplicity of the astroids concept and could be expanded upon with more ideas later, great!
Some things we should consider, First person or Third person?
If we use Player Vs. Player we need to incorporate networking
or if we use Artifical Inteligence we need a path finding system to avoid obsticables on their way towards the player to attack.
Lets remember to Keep it simple and not try to always go with the "Both" answer. cheers mates!
Swords are hard actually. It involves a fair amount of animation and suchlike.
Something like a tank fight in a maze or as already suggested an asteroids clone would be simpler.
Your first release should absolutely be single player too. Multi player is an order of magnitude harder and MMO another order of magnitude harder than that!