Support NeoGAF

[razu]

Anyone have good reading/articles/tips on AI in racing games? I'm working on something, but the AI is just miserably tough for me to figure out. I'm doing it in Game Maker (so GML), but the theory should be similar.

E: Also, is Unity a better option for an isometric racing game? I've been working in 2D, but it's nearly impossible to get a good sense of depth in 2D like that.

I wrote a super-simple racing AI system once. I did this:

Drove a few laps of the circuit manually.
Recorded the speed, direction and control state every 10 metres or so.
Created an average spline, with this data attached to each control point.
Wrote AI that tried to replicate the average run by starting on the spline and aiming at a point X metres further up the line. X depended on current speed.
It used the control data and speed/direction as hints as to whether to brake, accelerate, turn early..
The AI didn't cheat, all it did was generate control inputs and was run as normal.


I think it took about a day to write.

It was simple, but we had karts driving round tracks!

It could easily be extended to have several splines, to use for overtaking, etc..



Also, you can do anything in Unity. I'd at least suggest giving it a test.

 

[JulianImp]

Anyone have good reading/articles/tips on AI in racing games? I'm working on something, but the AI is just miserably tough for me to figure out. I'm doing it in Game Maker (so GML), but the theory should be similar.

E: Also, is Unity a better option for an isometric racing game? I've been working in 2D, but it's nearly impossible to get a good sense of depth in 2D like that.

If you wanted to turn it into a 3D game (even if the camera was set to orthographic), you should try Unity out. I think the pro version even has pathfinding support (but going pro will cost you quite a bit of money, at least when you're starting up).

Racing game AIs seem complicated to implement properly, since you want it to be smart enough to avoid glaringly obvious mistakes (ie: crashing into each other or running into hazards all the time) while being stupid enough to give players a shot at winning.

Here's a Gamasutra article I somehow managed to salvage (never deleting my browsing history does have its uses): http://www.gamasutra.com/view/feature/132313/the_pure_advantage_advanced_.php

Additionally, I think you should read up on some AI flocking implementations. They're quite useful for things such as following objects (in a racing AI's case, waypoints scattered accross the tracks or enemy cars), avoiding them and keeping the pack together.

One of the basic things you should do if you're going to have a multi-track point-based mode (such as Mario Kart's) is predetermine each AI's "preferred" position; otherwise, AIs might reach the goal in wildly varying positions, making things way too easy for the player. That's why most racing games always have a few AIs that play well and get the better positions most of the time, and other "bulk" AIs that pose little to no challenge to the player and keep doing badly.

@razu: That's a really interesting implementation (I love "fair" AIs). Is there any chance you could write a more detailed explanation of it (even if it's only slightly more detailed)? Overall, it sounds simple enough, but I still can't grasp how the recorded data was interpreted by the AI to decide which inputs to simulate. How did your system handle several racers on-track at once to prevent them from looking as if they were merely following each other?

 

[Ashodin]

Well I had a fun weekend in Guild Wars 2, but it's back to work this week

Hot Applesauce will be finished very soon I think.

 

[razu]

@razu: That's a really interesting implementation (I love "fair" AIs). Is there any chance you could write a more detailed explanation of it (even if it's only slightly more detailed)? Overall, it sounds simple enough, but I still can't grasp how the recorded data was interpreted by the AI to decide which inputs to simulate. How did your system handle several racers on-track at once to prevent them from looking as if they were merely following each other?

Thinking about it, we only had one test AI racer at a time. For demo mode, etc.. The studio was shut down before any more work could be done!

I think it was pretty basic in that if the point it was looking at had a higher speed, it would accelerate, lower, it would brake! But there was a bit of extra code to get the kart pointing the right way, as they did drift. So it didn't *just* steer right if the point was to the right, etc.. This was in 2003 I think, so details are a bit sketchy now!

The main takeaway point was that you can generate nice data from actually playing the game.

 

[JulianImp]

Thinking about it, we only had one test AI racer at a time. For demo mode, etc.. The studio was shut down before any more work could be done!

I think it was pretty basic in that if the point it was looking at had a higher speed, it would accelerate, lower, it would brake! But there was a bit of extra code to get the kart pointing the right way, as they did drift. So it didn't *just* steer right if the point was to the right, etc.. This was in 2003 I think, so details are a bit sketchy now!

The main takeaway point was that you can generate nice data from actually playing the game.

I guess that's right, you could play the game to generate some basic "recommended" path and then have the AI add some random offset to follow a slightly different path and flocking behaviors to avoid other racers. It'd be a good start with less strict requirements than actually having the AI recognize tracks on the fly.

 

[razu]

I guess that's right, you could play the game to generate some basic "recommended" path and then have the AI add some random offset to follow a slightly different path and flocking behaviors to avoid other racers. It'd be a good start with less strict requirements than actually having the AI recognize tracks on the fly.

In Colin McRae Rally 2 our AI programmer used and 'trained' neural nets.. That was a LOT of effort, but produced great results! Kind of the opposite approach to mine

 

[nicoga3000]

Some pretty killer info guys, thanks. I should have guessed that racing AI wasn't an easy task.

 

[usea]

Some pretty killer info guys, thanks. I should have guessed that racing AI wasn't an easy task.

http://books.google.com/books?id=1O...ficial intelligence for games driving&f=false

This book is really excellent for game AI, and here is the section on driving. Depending on the physics in your game, the degree of difficulty in composing actuators (accelerating, breaking, steering, etc) into goal-oriented motion can be more or less difficult.

 

[Ihateyouchris]

just wanted to throw in my two cents and let you know that any of the guys in a suit immediately made me think of braid. i know someone else mentioned that too so... yeah. if you're cool with it that's fine. personally, with as abundant as platformers are on the indie scene i'd do everything i can to make sure my platformer didn't remind people of another one.

as a side note, i really like your pixel art. the colors are great and there's a certain "chunkiness" to it that i really dig. hopefully, i'll one day get to make a game with really, really kick-ass pixel art

haha, I totally love the title.

And I agree with Mike in the characters. There are so many indie 2D platformers that you really want your character to stand out.

Talking about other characters, I the glass-less version somehow reminds me of Spelunky's main character. I think it's the combination of thos small pixel-y eyes and the bulbous nose. It probably isn't that problematic, though.

Nitpicking a bit, I think you should try to add a little more contrast between the nose and the rest of the face and between the beard and the shirt (dark brown on dark grey). That way they'd be easier to read when shown at 1x zoom.

Thanks for the feedback. I hadn't thought about Braid while making the suit version, so that's probably out. I might put in the suit as an item with a timey wimey ability though. Here's my current favorite:

In other news, I made "Bad Driver! +1" for ludum dare this past weekend. I ended up going to local meetup for the jam, and it was a blast! I highly recommend going to one if you get the chance.

 

[AusRoachman]

So ive been prototyping different themes / graphic ideas for awhile now for my iphone/ipad platformer and then i thought i might see what a rogue/fantasy type setting might look like ,ere is my concept

All art was done by oryx design labs and if i go ahead with it i will license it

I kinda of like the simplistic fantasy look, was planning on you playing as the Archer and travelling across lots of different style lands and dungeons .

I also have plans to put in a spelunky style random level generation and 1 death mechanic.

But gameplay wise it would be a traditional platformer with some shooting and lots of collecting.

What does everyone think ?

 

[JulianImp]

So ive been prototyping different themes / graphic ideas for awhile now for my iphone/ipad platformer and then i thought i might see what a rogue/fantasy type setting might look like ,ere is my concept

All art was done by oryx design labs and if i go ahead with it i will license it

I kinda of like the simplistic fantasy look, was planning on you playing as the Archer and travelling across lots of different style lands and dungeons .

I also have plans to put in a spelunky style random level generation and 1 death mechanic.

But gameplay wise it would be a traditional platformer with some shooting and lots of collecting.

What does everyone think ?

I absolutely love the micro pixel art aesthetic! The one complaint I'd have is that it all looks too similar to Zelda characters (mostly Link and the red bubbles), so you should try a slightly different approach to avoid legal implications and, at the very least, people randomly decrying your game as a Zelda knockoff like how they often call all low-res voxel games Minecraft clones.

I guess as far as mechanics go you could make it a left-to-right platformer rather than a top-to-bottom one like Spelunky to differentiate your game from it. If you could stick multiplayer co-op there, I'd love you even more (even if it's just local).

</rant>
Seriously, the world needs more local co-op games!
</rant>

Yesterday, I bought an Xperia Tipo for playtesting Quark Storm, and today I deployed the game and it ran fine, except for some GUI elements which had the wrong sizes. Its small 320x480 resolution will help me a lot in making sure the game looks fine across all screen sizes.

 

[_machine]

I really like the aesthetics and the concept, but why no paperdoll-esque equipment?

It's always my biggest complaint in the most recent kinda roguelike games like Dredmor or Hack, Slash and Loot; if there's gear it should also be visible and graphically different. I always have problem whoring loot if it doesn't have an effect on my character's look.

What are you using to create the game?

 

[Raide]

Looks like Realm of the Mad God in side-scrolling form. Could be pretty awesome.

 

[nicoga3000]

Looks like Realm of the Mad God in side-scrolling form. Could be pretty awesome.

It appears to be using the same resource pack they released for that challenge (I think it was one of those Game Jam's or weekend deals). Definitely interested in what gets put together!

 

[AusRoachman]

Looks like Realm of the Mad God in side-scrolling form. Could be pretty awesome.

I believe it is the same sprite pack

Available from here http://oryxdesignlab.com/sprites/

So many awesome sprites, but can only be use for non commercial games, however i have spoken to its creator and he has agreed to let me license them if/when i get to the stage of selling it.

It is currently built on cocos2d and plays like a traditional platformer (spikes, enemies, collectables e.t.c) however i am contemplating putting a more rogue like spin on it (randomised levels, 1 life similar to spelunky).

I envision it would playout abit like a spelunky level on sideways instead of downwards, but would also switch it up with auto scrolling levels, some that go up , some that go down e.t.c

Throw in other items like loot that gives random bonuses and lots of different enemies i envision it would be quiet cool.

I would love it to be a sort of lord of the rings vibe where each level has you meeting a different race (elves,dwarfes,ents e.t.c)

Anyway just an idea atm but so far its my frontrunner for my next prototype.

 

[Raide]

I believe it is the same sprite pack

Available from here http://oryxdesignlab.com/sprites/

So many awesome sprites, but can only be use for non commercial games, however i have spoken to its creator and he has agreed to let me license them if/when i get to the stage of selling it.

It is currently built on cocos2d and plays like a traditional platformer (spikes, enemies, collectables e.t.c) however i am contemplating putting a more rogue like spin on it (randomised levels, 1 life similar to spelunky).

I envision it would playout abit like a spelunky level on sideways instead of downwards, but would also switch it up with auto scrolling levels, some that go up , some that go down e.t.c

Throw in other items like loot that gives random bonuses and lots of different enemies i envision it would be quiet cool.

I would love it to be a sort of lord of the rings vibe where each level has you meeting a different race (elves,dwarfes,ents e.t.c)

Anyway just an idea atm but so far its my frontrunner for my next prototype.

Looks fun. Will keep an eye on it.

 

[JulianImp]

In other news, I made "Bad Driver! +1" for ludum dare this past weekend. I ended up going to local meetup for the jam, and it was a blast! I highly recommend going to one if you get the chance.

It's a hilarious game! Before playing it, I thought the bonus indicators were persistent or something... that would've been interesting as well, by reducing the number of pop-ups per trick but eventually filling the whole screen with them, therefore rendering the player unable to see what he's doing (that'd make adding audio a priority, though).

The slippery movements actually added to the game's wackiness. IT was really fun trying to hit something but failing miserably and crashing.

By the way, I experienced a small bug when I crashed while giant-sized, and the car just kept flying upwards until I was so high up the level stopped being rendered. Still, I saw it as more of a feature than a bug, and just laughed for a while before finally exiting the game.

 

[Heart of Black]

I wanted to do Ludum Dare this weekend, but had a shoot for my wife's/friend's web series + an event to go to AND no one set up a meetup in New York anyway.

Global Game Jam is coming up in Jan, so that looks good.


Right now I am working with a friend (programmer) to figure out the intricacies of the Playstation Mobile Development suite. As soon as we get the game (prototyped first in Game Maker) running I'll start to chronicle its development.

 

[JulianImp]

I've just read on Twitter that Cook, Serve, Delicious! was finally released for iOS.

I think it was developed by some fellow GAFfers (I think chubigans), wasn't it? Congratulations to the team for the new release!

 

[Roquentin]

I've just read on Twitter that Cook, Serve, Delicious! was finally released for iOS.

I think it was developed by some fellow GAFfers (I think chubigans), wasn't it? Congratulations to the team for the new release!

Yup. From previous page:

Wrapped up my first iOS game and released it. Man, I LOVE the iPad platform. I really hope the game does well for the simple fact that I just wanna make iPad games forever.

Here it is- https://itunes.apple.com/us/app/cook-serve-delicious!/id582153229?mt=8

 

[Genji]

Been working on the combat UI for my game after switching over from iOS -> PC. Have tried out various ways of showing the enemy health. Tried rts style health bars over each enemy and also smoke from damaged enemies. Both felt a bit too busy so I settled on a color gradient for enemy hull/shield strength on the top right of the hud along with an indicator for where the targeted enemy is in relation to the player when the enemy is off-screen. Still need a good indicator for the targeted enemy when it's on-screen.

 

[Dynamite Shikoku]

So ive been prototyping different themes / graphic ideas for awhile now for my iphone/ipad platformer and then i thought i might see what a rogue/fantasy type setting might look like ,ere is my concept

All art was done by oryx design labs and if i go ahead with it i will license it

I kinda of like the simplistic fantasy look, was planning on you playing as the Archer and travelling across lots of different style lands and dungeons .

I also have plans to put in a spelunky style random level generation and 1 death mechanic.

But gameplay wise it would be a traditional platformer with some shooting and lots of collecting.

What does everyone think ?

I love the look of that

 

[usea]

Genji that looks awesome.

All of you guys are a really great motivator. I want to make stuff as cool as what you are making.

"guys" being a gender-neutral colloquialism.

 

[charsace]

razu, I'm still thinking of making Chop Her, Mike!.

 

[bumpkin]

Admittedly, all of my game dev exploits have been on hold as of late. After several weeks of spinning my tires and not making any progress, I decided to direct my attention to updates I've been meaning to make to an iPhone App I published last year. It's nothing gaming related, so not worth calling out or sharing. But once I get those updates done and publish the new version, perhaps I'll return to my game engine project.

Perhaps.

If I didn't have a near-defiant love affair with playing games with a keyboard or controller, I'd give serious thought to making something for iOS.

 

[Ashodin]

Progress is a lot slower on creating the powers than I thought for Mr. Wave. It's a lot of clerical work fixing up the controls (thankfully, it's a lot of cut and paste).

Will be playable soon!

 

[Ranger X]

So ive been prototyping different themes / graphic ideas for awhile now for my iphone/ipad platformer and then i thought i might see what a rogue/fantasy type setting might look like ,ere is my concept

All art was done by oryx design labs and if i go ahead with it i will license it

I kinda of like the simplistic fantasy look, was planning on you playing as the Archer and travelling across lots of different style lands and dungeons .

I also have plans to put in a spelunky style random level generation and 1 death mechanic.

But gameplay wise it would be a traditional platformer with some shooting and lots of collecting.

What does everyone think ?

I love that style. I feel like shooting could be awesome. I wouldn't go the "overbleeding" cliché route but make the shooting satisfying and most importantly, the platforming.
Is this "adventure style" where its one open world (Metroid) and you progress depending the abilities you collected? I feel like I would want to swim in that screen shot and then dig in order to fall in that brown corridor.

 

[JulianImp]

Genji: That looks slick! Perhaps you could mark onscreen targeted enemies with a contrasting colored outline?


bumpkin: I think long development periods can put quite a stain on your ability to get things done.

Working on my own on Quark Storm has been quite a ride, with lots of moments where I was unable to get many things done and started wondering if I should've just dropped it and begun on something new.

Once I finally began to speak up and make the game's development public, however, not only did I get words of encouragement from people who like what I was working on, but I also commited to that people. It felt awesome seeing them playing and enjoying something I made, and was a great positive feedback loop to make me push onwards. Before I was doing Quark Storm for myself, but now I can't let them down!


Ashodin: If you're copy-pasting code often, you should consider making a more robust implementation; for example, interfaces or an abstract "power" class which each power inherits and implements its own way.

Projectile-firing powers could also have a base implementation, with variables such as projectile type, speed, graphics and whether they're shot constantly as long as you hold the button down (with a variable to control how long it takes between shots, of course!) or a single shot per button press. Things like this will save you a lot of time in the long run with minimum ammounts of planning required so you know what kinds of variables you want your base clases to have.


That rogue concept is making me want to boot up GameMaker once again to build something. Pixel art like that would be hard to implement in Unity without specialized add-ons.

I'd love to see a sidescroller that looked like that, but the sprites got me thinking on a local co-op roguelike:

• Semi turn-based gameplay
• All characters and objects (yes, even projectiles) have an "action bar" that fills up over time
• When the action bar's full, the character does whichever action was buffered last
• Actions would include:
• Standing still
• Moving to an adjacent tile
• Using an item (out of one or two slots, since it'd be hard to look through a menu in time)
• Weapon attack (can be directed)
• Cast magic (can be directed)
• Use object? (could be simplified to merely stepping on the object's tile)

I guess that could've been done before, but it got me really interested. Having several people playing at once would lead to hilarious situations, such as a fast acting player beating the rest to the loot, or having to coordinate movement to make sure they don't block each other's paths or projectiles.

God, I can't stop thinking about this, I love the concept too much! I must resist the temptation and get back to finishing level 1-3 for Quark Storm! D:

 

[nicoga3000]

Interesting idea! Especially the idea of everything having action bars.

My gripe with ideas is my lack of resources. If I had a huge pool of graphics to choose from, my life would be complete.

 

[Ashodin]

Ashodin: If you're copy-pasting code often, you should consider making a more robust implementation; for example, interfaces or an abstract "power" class which each power inherits and implements its own way.

Projectile-firing powers could also have a base implementation, with variables such as projectile type, speed, graphics and whether they're shot constantly as long as you hold the button down (with a variable to control how long it takes between shots, of course!) or a single shot per button press. Things like this will save you a lot of time in the long run with minimum ammounts of planning required so you know what kinds of variables you want your base clases to have.

Actually this is how I have it set up. All powers will be single shot attacks, yes. This is all me generating the code from scratch, but once it's generated, all I really have to do in the future (ie new powers) is copy/paste the code, change some parameters, then add new graphics. Everything right now is laying the groundwork to make it easy in the future.

 

[JulianImp]

Actually this is how I have it set up. All powers will be single shot attacks, yes. This is all me generating the code from scratch, but once it's generated, all I really have to do in the future (ie new powers) is copy/paste the code, change some parameters, then add new graphics. Everything right now is laying the groundwork to make it easy in the future.

Well, copy-pasting code could be a bad practice. It'd be best if you had something like a generic projectile class, like (in Unity C#):

public class Projectile {
    public enum Element {
        Fire,
        Ice,
        Lightning
    }

    private float     speed;
    private Element   element;
    private int       baseDamage;
    private Texture2D texture;

    static private bool        lutsUninitialized = true;

    static private float[]     projectileSpeeds;
    static private int[]       projectileBaseDamages;
    static private Texture2D[] projectileTextures;
    
    protected Projectile (Element element) {
        if (lutsUninitialized) InitializeLuts();

        int elementIndex = (int)element;

        this.speed = projectileSpeeds[elementIndex];
        this.element = element;
        this.baseDamage = projectileBaseDamages[elementIndex];
        this.texture = projectileTextures[elementIndex];
    }

    static private void InitializeLuts() {
        lutsUninitialized = false;

        int maximumElementIndex = ((int)Element.Lightning) + 1;

        projectileSpeeds = new float[maximumElementIndex];
        projectileSpeeds[Element.Fire]      = 0.25f;
        projectileSpeeds[Element.Ice]       = 0.5f;
        projectileSpeeds[Element.Lightning] = 0.75f;

        projectileBaseDamages = new int[maximumElementIndex];
        projectileBaseDamages[Element.Fire]      = 6;
        projectileBaseDamages[Element.Ice]       = 3;
        projectileBaseDamages[Element.Lightning] = 2;

        projectileTextures = new Texture2D[maximumElementIndex];
        projectileTextures [Element.Fire]      = Resources.Load("FireParticle");
        projectileTextures [Element.Ice]       = Resources.Load("IceParticle");
        projectileTextures [Element.Lightning] = Resources.Load("LighningParticle");
    }

    //Add code for movement and texture animation here
}

With that in place, firing a lightning shot could be as easy as calling "new Projectile(Projectile.Element.Lightning);". All movement, animation and collision checking would be written once, instead of in all three shot types' classes.

 

[FormallyMegadrive]

The above is good advice. When you find yourself copy-pasting a lot of similar code, always look into refactoring. I'd even look into having separate Projectile-inherited classes for Lightning and so on, with a generic Update() method being called each loop.

 

[Ashodin]

I guess I should add that I'm working on Construct 2 framework. Not exactly "coding" as more "organizing the code" heh.

 

[FormallyMegadrive]

Well yeah. From memory Construct uses an event-style thing similar to the Clickteam products, and not code in any form. Bummer.

 

[Tankshell]

Hey guys,

My app has been submitted to the appstore for review (hopefully approved before end of this week!). I have been designing the promotional website to along with my game (Ultra Super Firework Planets: In Space)... the site is not actually live yet but here is a mockup screenshot of it:

Do you think it looks ok? Will be creating the OT for my game soon too, hope to see some of you over there supporting fellow GAF indie dev efforts (well, at least those of you with iOS devices!) =0)

 

[Ashodin]

I'm not too sure you can create your own OTs for your own stuff. Self-promotion on GAF is frowned upon.

 

[Tankshell]

I'm not too sure you can create your own OTs for your own stuff. Self-promotion on GAF is frowned upon.

Orly? That sucks. Oh well I won't bother then =0) Am sure I have seen other indie devs on GAF create their own before though.

Regardless, any thoughts on website design? Being my first indie game I'm not sure what else to put on there, I think I have all the basics covered: screenshots, information, support section.

Only other thing I want to do is a trailer of some sort.

 

[Feep]

I'm not too sure you can create your own OTs for your own stuff. Self-promotion on GAF is frowned upon.

Yeah, I would get one of us or a friend to do it for you.

But also, the design and UI work looks hyper slick. Looking forward to trying it!

 

[Ashodin]

Yeah we would totally promote your game for you. And it's why you see a lot of indie devs post their game in this thread for exposure (being that someone else needs to post the thread for their game).

 

[fin]

GITHUB posted the game-off winners.

https://github.com/blog/1337-github-game-off-winners

My submission never won, runner-up or honorable mention. Kinda shitty but oh well. Not sure if I'm going to keep working on it. I need practice with doing art so I think I'll keep it going a bit longer. Got it working on Android, so maybe see how far I can go with it till Ouya launch. If anyone missed it you can play it here:

http://goo.gl/OuPnr

 

[Genji]

Genji that looks awesome.

All of you guys are a really great motivator. I want to make stuff as cool as what you are making.

Thanks! I totally agree with you in that this thread and the contributors serve as a great motivator. Being able to see what others are working on, their ideas, their progress and, probably most importantly, their enthusiasm has helped push me to some late night dev sessions.

Genji: That looks slick! Perhaps you could mark onscreen targeted enemies with a contrasting colored outline?

Thanks, I'll give that a try, would be a clean way to indicate the target. If you decide to implement your rogue idea (which sounds very cool btw, would be hilarious to block a fellow player and have them die to mobs because he can't pass) using Unity, I've had good results using 2D Toolkit.

 

[Mario]

Orly? That sucks. Oh well I won't bother then =0) Am sure I have seen other indie devs on GAF create their own before though.

You can post in the iOS OT without fear.

Creating your own OT (for an iOS game no less!) will bring on the wrath of GAF.

 

[Mista Koo]

So I joined the latest Ludum Dare (Theme: You are the Villain) with this gem.. Super Goon.



Experience a day in the life of a goon. An art game with a "deep message."
(It only takes a minute to try "everything" in it.)

I'm happy to finally post a game(?) I made here, especially since I have been posting about wanting to make something since the start of the previous thread

 

[Tash]

Hey guys,

My app has been submitted to the appstore for review (hopefully approved before end of this week!). I have been designing the promotional website to along with my game (Ultra Super Firework Planets: In Space)... the site is not actually live yet but here is a mockup screenshot of it:



Do you think it looks ok? Will be creating the OT for my game soon too, hope to see some of you over there supporting fellow GAF indie dev efforts (well, at least those of you with iOS devices!) =0)

REALLY love the look of the website and yeah, in general it's a bad idea to pitch your own stuff
Remember to add a presskit for download to your site. Good for journos.

 

[JulianImp]

Thanks, I'll give that a try, would be a clean way to indicate the target. If you decide to implement your rogue idea (which sounds very cool btw, would be hilarious to block a fellow player and have them die to mobs because he can't pass) using Unity, I've had good results using 2D Toolkit.

I'll see if I can come up with a custom framework, using texture atlases and a rudimentary animation system (since it's not like 8x8 px graphics can move much). Flash would be awful because of no gamepad support (good luck getting several players on a single keyboard), and I'd need to buy a GameMaker license to do anything worth my while.

In Unity, I'll probably have to relinquish its collision system and make a tile-based map system for movement and graphics, probably using a low, Game Boy Advance-like resolution with the ability to scale it up in integer increments (210x160, 420x320, 630x480, etcetera) so there're no awful sub-pixels being rendered.

I might start it as a short side-project, at least to see if I can get a prototype done before going back to Quark Storm. The good thing is I'm somewhat competent at drawing pixel art, so I'll be able to make the game, its graphics and music (as long as I stick to simple chiptunes).

 

[Miutsu]

Sup guys and gals,

I've have submitted my game to the Google Play market last night and now its up:

https://play.google.com/store/apps/details?id=com.fe.segame

It's a tile puzzle game with some beats thrown in the mix, the gameplay revolves mostly around finding the right path in a grid of tiles in order touch them all without repeating tiles or reaching a dead end, very much like some of the puzzles found in some of the GB Zeldas.

The game was made for the Square Enix Latin America contest, it lacks a little polish in some parts but development had to stop at some point for external reasons and we decided to release it as is, the game is free right now and will be (maybe, I'm not sure yet) paid after some time.

It was made using LibGDX (awesome free lib guys) which currently supports Android, HTML5, Desktop and soon iOS.

If you try it tell us what you think

 

[razu]

razu, I'm still thinking of making Chop Her, Mike!.

Go for it!

Make it so that when you slice the ladies you see a cross-section of their innards!

I'd play THAT game!

 

[darkblade77]

Thought I'd drop by in here to get some pointers...

I'm building a 2D fighting game engine in Unity, and I'm having trouble figuring out whether I want to dip into component-based design for some of the core elements(for example, I have to roll my own collision and pseudo-physics). This is mostly because I don't feel like I have a comfortable understanding on how components in Unity really work.

I know how to use existing components; for example my engine relies on ex2D for handling images right now.

What I'm worried about is mostly performance. Fighting games need to perform consistently with ideally no pauses or hiccups during gameplay, and they need to run on fixed timestep. Setting the game to run like this isn't a problem, but I have no idea if a traditional OOP design or a component-entity design would ensure the least amount of burden when anything that updates is only calling FixedUpdate.

I have an OOP-centric prototype working right now... I'd like to build a component-driven prototype to test it against, but I'm also not sure what would be good criteria for a performance test.

Any suggestions would be appreciated.

EDIT: Well, talking about it made me think of a few answers, so I'm going to try some stuff out. But any knowledge on the subject is welcome!

Also, and I guess this is kind of a shot in the wind, if there are any artists/animators that would be willing to let me use their assets just for testing(so if I reach a point where I can show things off, it won't be copyright infringement city like it is now), please let me know.

 

[Tash]

Sup guys and gals,

I've have submitted my game to the Google Play market last night and now its up:

https://play.google.com/store/apps/details?id=com.fe.segame



It's a tile puzzle game with some beats thrown in the mix, the gameplay revolves mostly around finding the right path in a grid of tiles in order touch them all without repeating tiles or reaching a dead end, very much like some of the puzzles found in some of the GB Zeldas.

The game was made for the Square Enix Latin America contest, it lacks a little polish in some parts but development had to stop at some point for external reasons and we decided to release it as is, the game is free right now and will be (maybe, I'm not sure yet) paid after some time.

It was made using LibGDX (awesome free lib guys) which currently supports Android, HTML5, Desktop and soon iOS.

If you try it tell us what you think

Good luck, I like how slick it looks

 

[Miutsu]

Good luck, I like how slick it looks

Thanks! It doesn't have many assets but the ones that are there I made them myself what sucks is that I only made them for one target resolution and in anything other than that it scales the assets, which still manage to look good from the devices I've used to test but yeah, next time I'll try to support more target res.