Development Direction for Game

Anything about development not directly related to libgdx, e.g. OpenGL, Android APIs etc.

Development Direction for Game

Postby DRing » Thu Feb 08, 2018 6:37 am

So this is more of a discussion then a specific question, Im starting a new project and before I go to far into it I wanted to get some opinions on the approach to take. So first the game is going to be 2d side view fighter type game, think street fighter for simplicity and not going into too much detail. With this I have a few things that not quite sure which method to take. Here are my few questions and just looking for general approach on these:

1. Is box2d a good idea? I know there is some benefit to using it, but for a fighter type game im going to have special moves wondering if it will end up being too much of a hassle to get the movement or if I should just handle movement and collision without it. I will have platforms that they can jump thru from below as well.

2. Game Controls - first thought is to do an onscreen joystick and buttons, but personally those never feel great. Wondering if someone has other ideas for movement and attacking.

3. AI for NPCs - Ive dealt with simple pathfinding and reaction type before, but this will be a little more involved, wondering if there is already an approach, library or something for this. I havent been able to find anything so far.

Again just looking for some opinions on these topics
DRing
 
Posts: 104
Joined: Thu Oct 20, 2011 7:42 pm

Re: Development Direction for Game

Postby RafaSKB » Thu Feb 08, 2018 11:41 am

1. Box2D: Unless you want realistic physics simulation and have complex worlds, procedural platforms, slopes and whatnot, I don't recommend using it because it can get complicated fast, and I believe you're aiming for something simple. Manual collision detection for what you described shouldn't be too difficult to do.
A simple example of how Box2D can be a problem is the fact bodies can't be resized after they're created, unless they're destroyed and created again with the new size. If your character ducks, you will either have to recreate the body or have their bodies made out of two segments, top and bottom, and have the top one disabled until they stand again.

2. Controls: I don't have much experience with mobile controls, but I've seen some games that do a sort of onscreen joystick, however they're not fixed buttons on the screen, but rather actions that respond to touch gestures, so if you drag your finger from point A to point B in the left area of the screen, going left, the character moves left, without really having to press that exact button position.
That being said, I don't play many mobile games at all and I'm not sure if that would work for a fighting game.

3. AI: There's gdx-ai out there, which is very good. Might need some adaptation for platformer games though (or not, never used it). If not, there are several tutorials out there that explain how to do platformer pathfinding and other things.
Typing Label - Label that appears as if it was being typed in real time. https://github.com/rafaskb/typing-label/
Grashers - A trash-themed roguelike in space, soon on Steam. - http://www.grashers.com
RafaSKB
 
Posts: 122
Joined: Sat Aug 01, 2015 5:14 am
Location: São Paulo, Brazil

Re: Development Direction for Game

Postby Magnesus » Thu Feb 08, 2018 1:18 pm

DRing wrote:1. Is box2d a good idea? I know there is some benefit to using it, but for a fighter type game im going to have special moves wondering if it will end up being too much of a hassle to get the movement or if I should just handle movement and collision without it. I will have platforms that they can jump thru from below as well.


IMHO definitely. RafaSKB described the common problems you might encounter - but keep in mind that recreating and creating bodies is very fast, so it's not a big problem if you need to do it. I would definitely make the characters from more than one part though and moved or disabled the top part for ducking. For attacks I would create additional special bodies that would detect what the attack hit. (If there will be any platforming read about unicycle character in box2d.)

2. Game Controls - first thought is to do an onscreen joystick and buttons, but personally those never feel great. Wondering if someone has other ideas for movement and attacking.


People hate onscreen controls, but it's hard to make anything else work for such a game. Swiping over the characters is also bad since you don't see what is happening. Maybe check how similar, popular games do it on mobile? Also consider making the game turn based - it could be fun on mobile.

3. AI for NPCs - Ive dealt with simple pathfinding and reaction type before, but this will be a little more involved, wondering if there is already an approach, library or something for this. I havent been able to find anything so far.


As RafaSKB said - check gdx-ai first.

Look also in Spine2D or something similar, unless you are going for pixel art. Good luck.
Magnesus
 
Posts: 1632
Joined: Sun Sep 25, 2011 3:50 pm

Re: Development Direction for Game

Postby RafaSKB » Thu Feb 08, 2018 1:46 pm

Magnesus wrote:Look also in Spine2D or something similar, unless you are going for pixel art. Good luck.


Good call! Spine is perfect for this kind of game. :)
Typing Label - Label that appears as if it was being typed in real time. https://github.com/rafaskb/typing-label/
Grashers - A trash-themed roguelike in space, soon on Steam. - http://www.grashers.com
RafaSKB
 
Posts: 122
Joined: Sat Aug 01, 2015 5:14 am
Location: São Paulo, Brazil

Re: Development Direction for Game

Postby DRing » Thu Feb 08, 2018 4:02 pm

1.Well guess Ill have to do more research on whether or not to use Box2d, I have a feeling the movement is going to be a pain and if I can handle collision on my own relatively easily then Ill probably do that so I can make the movement feel the way I want.
Magnesus wrote:If there will be any platforming read about unicycle character in box2d.

Do you have a link to that info

2. Yeah Im not a huge fan of on screen controls either, but currently cant think or havent seen a better method for this. Turn based isnt really an option either

3. Ill have to look more into gdx-ai, thanks for pointing me into that direction, guessing Ill use the movement and pathfinding in conjunction with behavior trees for their attacks.


Magnesus wrote:Look also in Spine2D or something similar, unless you are going for pixel art. Good luck.

I have seen this once before and will need to look more into it. Looks like it is worth the $70 to get started
DRing
 
Posts: 104
Joined: Thu Oct 20, 2011 7:42 pm

Re: Development Direction for Game

Postby DRing » Sun Feb 11, 2018 3:10 am

So going back to whether to use box 2d or not. What about using Actors and a stage to do the movements. Libgdx stage and actors have a collision system too correct?

Thoughts? Pros/Cons?
DRing
 
Posts: 104
Joined: Thu Oct 20, 2011 7:42 pm

Re: Development Direction for Game

Postby RafaSKB » Sun Feb 11, 2018 1:11 pm

Scene2D was made with UI in mind, not really suited to hold game elements. I honestly think you will have a hard time if you choose Scene2D for that.
Typing Label - Label that appears as if it was being typed in real time. https://github.com/rafaskb/typing-label/
Grashers - A trash-themed roguelike in space, soon on Steam. - http://www.grashers.com
RafaSKB
 
Posts: 122
Joined: Sat Aug 01, 2015 5:14 am
Location: São Paulo, Brazil


Return to General Development

Who is online

Users browsing this forum: No registered users and 1 guest

cron