[jME Game] Fighting the white bull

Ok, now we have finally decided for a story. Now it's time to setup a project and lay the foundations for the game.

To be honest, i have never setup a game project.



First of all, we need a place where we can put all of our work. This can be found here: http://code.google.com/p/proxywars/



What are the next steps? For the code i would strongly suggest, that we start with a proper environment, where we can build the sources from the beginning.



Here's my list:

  1. Setup a build system (having no code yet, this will be a very interesting job ;))
  2. Write some sketch for the story
  3. Write a game design document
  4. Define the rules for the project
  5. Create a roadmap



    Please add items to this list.



    Also In this thread i would like to see people volunteer for any task that we identify here. I will create issues in the tracker and assign team members to them.



    Ok, then, let the party begin  :smiley:

I have checked in a prototype RTS engine. It's not perfect, maybe its even wrong or bad, but its a good starting point for discussions. Take a look  at http://code.google.com/p/proxywars/source/browse/

Hello,



just to say that I would like to volunteer for whatever task on this project, just one problem: my experience in programming (java, jME), graphics, storytelling … is pretty low  :(  :(. so I fear that I wont be able to do a lot.



If there is something I can do (always possible), please tell me!



I think this is a great project and I do wish you the best of luck!

One of the first bits of advice given to new people coming to jME and game programming in general seems to be: stop planning and brainstorming so much and just get something done… and very many game projects in general are said to fail for a similar reason - people come together… talk visions, what they want to do, throw ideas etc… and then they fall apart as no-one really codes anything.



I thought quite long before posting this… as I am not a part of this project and really have no say. But I am a part of the community and wish you all the best and wish you would succeed in your task… I really think having another prolonged poll and vision contest is a bad idea… you should start on the actual game already… unless you want it to die before it even takes off the ground.

I agree with Mindgamer. We/You can talk for hours, days… discussing about everything without doing anything, but how longer we do so, how longer it will take to really get something done and how bigger the chance that nothing will get accomplished.



We/You have already a story and a genre (RTS), Isn' it possible to start building an RTS and see how it grows? There will always be work that has to be redone and having 'a vision' to early can only obstruct developpement (in my opinion of course).



If this project is going to be big, there wil be time enough to discuss.

Mindgamer said:

One of the first bits of advice given to new people coming to jME and game programming in general seems to be: stop planning and brainstorming so much and just get something done..


I had the same concerns the whole week. So i renamed this topic to "Fighting the white bull". Hemingway called blank white papers the "white bull". Starting a new novel was always fighting with the white bull.

I'd say there are three main areas, that we need to work on:
1. Adminstration (roles, responsibilites, organization, etc.)
2. Technology (frameworks, libraries, tools, build, etc.)
3. Content (graphics, sounds, effects, models, story, etc.)

Each of this areas needs at least one team member, that declares himself responsible for that. I would volunteer for the first (not that i like it so much, but someone has to do the job)

But first of all, join the team at http://code.google.com/p/proxywars/

how to join the team…i visit the site specified …

Just send an email with your jME username and your Google account email address to ablochberger@gmail.com

thanks…i just sent the mail to the id specified by you…

Here are the next steps for each new member:


  1. Join the team, as described here
  2. Add rows to the skills table here

I’ve already started to create issues and assigned team members. So please watch the list of issues

thats cool, i wish i had more time to get involved with the project.

Nice, we can work off of that. Maybe load in a model, and get some physics working. BTW, we are(or will be) using JME 2.x, right?

I thought more about some factories. The framework is independent of the rendering system and uses factories to create the entities of the world.

So for the concrete game, you add some factories to the framework, add some listeners and controllers, that's the way i was thinking of

zagibu said:

But where are the comments?


Well, yes, this is true. But this code was only a shoot from the hip. But did you notice all the nice javadoc in IUnit? Of course, javadoc is essential, and i'll add this asap.

Jedimace1 said:

Nice, we can work off of that. Maybe load in a model, and get some physics working. BTW, we are(or will be) using JME 2.x, right?


I would say we are going for jME 2
The idea behind this interface based design is, that we can separate the model (here its the RTS engine) from the view (which will be jME)
So we can create a Node descendant, that implements IUnit. The framework only deals with IUnit, so it doesn't matter, what concrete implementation we have. But yet we have to analyze this model and find its weaknesses. Event-handling may be one of those weaknesses.

Core-Dump said:

thats cool, i wish i had more time to get involved with the project.

Time is only an illusion :P

That's a nice start. But you have to implement 3D rendering soon. In my experience it is much harder to implement a gameplay framework separate from a rendering system.



I suggest you have a "World" class where you can add & remove entities. As you add them, their Spatial(s) will be added onto the rootNode which you can render.

See this class: http://code.google.com/p/radakan/source/browse/trunk/radakan/src/com/radakan/game/world/World.java

The "World" can be an implementation of the "Framework" class you have.

I'm not sure if this is the right place (perhaps better in the google-page) or if I'm the right person to say but I wander if it would be possible to arrange a chat-session with all the/as many as possible project members. It will probabely be impossible to get everybody together on the same moment but I think it would really help alot to start things up.


just come to #jme on freenode always when you are online and you can chat to your heart's content. Might even find new members to the project

I'll try to go to #jme as often as possible