Quote:Original post by Firestorm ZERO
A group of my friends and I want to make a 2D shooting game with network capabilities. (like those top-down shooters or like Gradius with network abilities).

If you're just starting then a networked shooter is probably not a good idea. Scrolling shooters like R-Type and Gradius require *exact* movement and collision detection, and typically are one-hit-kills. This makes them *very* hard to make work over the internet. It'd be harder than a typical FPS or racing game (which tend to do all sorts of fudges and faking), and most of the information on the 'net doesn't apply.

You'd be much better off making a scrolling shooter with only local multiplayer (ie. two people on one keyboard). You're more likely to finish something playable that way.

LWJGL (or Slick) is a good choice though.

Quote:Original post by OrangyTang
If you're just starting then a networked shooter is probably not a good idea. Scrolling shooters like R-Type and Gradius require *exact* movement and collision detection, and typically are one-hit-kills. This makes them *very* hard to make work over the internet. It'd be harder than a typical FPS or racing game (which tend to do all sorts of fudges and faking), and most of the information on the 'net doesn't apply.

You'd be much better off making a scrolling shooter with only local multiplayer (ie. two people on one keyboard). You're more likely to finish something playable that way.

LWJGL (or Slick) is a good choice though.

I agree that a networked game for first timers is quiet a large challange but I will disagree that an FPS or Racing Game would be easier as those tend to be 3D and since you are suggesting they use LWJGL then I would highly advise against a 3D game. Mainly because you will have to make an engine on top of LWJGL that handles, rendering, device states, meshes, resource loading and caching, projections, tranformations, culling (most likely implementing an octree), animations, special effects and shaders (still have to do this with 2D), scene graph, and of course on top of all of this you will have to worry about the game play and gui interface. Then you also probably want networking for your FPS not so much for racing game. Though the racing game you'll probably have to fudge things to get the car physics to be just right. The wheels to turn just right, flipping if you go to fast over a bump, wheel and tire traction to make it appear as if the car is skidding. Definitely not something I would start out with as my first game especially on top of LWJGL. You might be able to do it with jMonkeyEngine but even then you need to know all the features a 3D engine provides in order to properly utilize them.

@Firestorm ZERO
My advice if you want to get something done and in a reasonable amount of time. Lets say a 2D Shooter. So you have a bunch of design options and some easily reachable requirements:

Requirements:
- 2D graphics
- PNG images w/transparency
- 16x16 image size
- 32x32 image size
- Sprite Sheets (for animation)
- Backgrounds ?x480 (if side scrolling) 640x? (if vertical scroller)
- Parallax Scrolling

Design Choices:
- Number of choosable ships
- Number of enemy ships
- Timed levels (or not)
- Scoring (points per kill/points per mission completed)
- Controls (movement, fire, switch weapons, special weapons)
- Multiple Weapons (or not)
- Gold per missions
- Warehouse if desired.

These I believe are reasonable for a first time project. Just remember that programming something you have never programmed before expect to add an additional hour for every hour you estimate it will take.

Technologies I suggest:

Java 2D (Golden-T Game Engine is ok) - yes I know you want something cool and spiffy and you want to add all those shader effects, but for a first project start simple.
java.net For the networking I suggest ironing out exactly what data absolutely needs to be sent across the connection, send only what is neccessary to those clients that need to know about it. There is a lot to making a game work across a network, latency, unexpected interruptions, security (encryption), packet loss, synchronization, message passing methods (pushing, polling), message queues, etc... So planning things out is key.

Thats really all you need and you won't have to work around compiling and running with added libraries and runtime-libraries which can be a pain if you've never used them before. That and you probably won't have to learn too much to get started on the programming since you said that you all have Java experience btw use swing.

Hope you take my advise, and good luck to you all!

Quote:Original post by 5MinuteGaming

Quote:Original post by OrangyTang
If you're just starting then a networked shooter is probably not a good idea. Scrolling shooters like R-Type and Gradius require *exact* movement and collision detection, and typically are one-hit-kills. This makes them *very* hard to make work over the internet. It'd be harder than a typical FPS or racing game (which tend to do all sorts of fudges and faking), and most of the information on the 'net doesn't apply.

You'd be much better off making a scrolling shooter with only local multiplayer (ie. two people on one keyboard). You're more likely to finish something playable that way.

LWJGL (or Slick) is a good choice though.

I agree that a networked game for first timers is quiet a large challange but I will disagree that an FPS or Racing Game would be easier as those tend to be 3D and since you are suggesting they use LWJGL then I would highly advise against a 3D game. Mainly because you will have to make an engine on top of LWJGL that handles, rendering, device states, meshes, resource loading and caching, projections, tranformations, culling (most likely implementing an octree), animations, special effects and shaders (still have to do this with 2D), scene graph, and of course on top of all of this you will have to worry about the game play and gui interface. Then you also probably want networking for your FPS not so much for racing game. Though the racing game you'll probably have to fudge things to get the car physics to be just right. The wheels to turn just right, flipping if you go to fast over a bump, wheel and tire traction to make it appear as if the car is skidding. Definitely not something I would start out with as my first game especially on top of LWJGL. You might be able to do it with jMonkeyEngine but even then you need to know all the features a 3D engine provides in order to properly utilize them.

I'm not refering to the 3d aspect - obviously 3d is somewhat more tricky than 2d - I'm talking about the networking aspect. Networking in a fps is basically a solved problem - basic client-server architecture, with server doing all the authentication and client-side prediction methods. It might not be easy to program but theres *loads* of information about this on the internet.

Try using this approach for a scrolling shooter like Graidus and it'll fail miserably. The sheer number of bullets and enemies to dodge and the accuracy required means that regular prediction methods won't be accurate enough and you'll get lots of frustrating and unfair deaths. You'd need a hybrid system taking the best bits of client/server and distributed simulation to make it play well.

Hey,

@OrangyTang

Quote:Original post by OrangyTang
I'm not refering to the 3d aspect - obviously 3d is somewhat more tricky than 2d - I'm talking about the networking aspect. Networking in a fps is basically a solved problem - basic client-server architecture, with server doing all the authentication and client-side prediction methods. It might not be easy to program but theres *loads* of information about this on the internet.

Try using this approach for a scrolling shooter like Graidus and it'll fail miserably. The sheer number of bullets and enemies to dodge and the accuracy required means that regular prediction methods won't be accurate enough and you'll get lots of frustrating and unfair deaths. You'd need a hybrid system taking the best bits of client/server and distributed simulation to make it play well.

Sorry for any little misinterpretation on my part! Just real quick, I'd like to say that I agree with you that there are many resources about FPS games and networking. And that integrating networking within a 2D shooter will be a challenge for FirestormZERO and team.

@FirestormZERO
Very cool! I'd be interested in knowing how things go with Slick as I haven't had the opportunity to use it. I'd like to know specifically how easy/hard integrating networking with Slick is since I've tried it for Golden-T and it was a nightmare.