LÖVE

Usually I'm very open about everything I'm working on, but this one I wanna kind of keep secret. It'll be my biggest project ever!! It'll be awesome. I'll be using LöveFrames for the UI(It's insanely awesome, if you haven't already I suggest you check it out!!!!) and probably AnAL for sprites - Though I don't know if I'll be using sprites.

I think it'll take me all summer to get a working alpha out, but development should go on until July 2014 because of the way the game is made and how I can add content. I hope LÖVE stays around for that long!! Maybe, if the versions of LÖVE add more that I can add, development could last longer!! I am sort of looking for people to design characters, but I'm holding back on it for now because of a few reasons:

1) I don't want to put people under pressure - Like if they take a long time to design something for no reason other than laziness and I tell them I need the image sooner they might get angry

2) I need to plan out the game more

3) I don't want to keep people back from their own IRL stuff e.g. Schoolwork.

4) My friend might be able to do it!

But yeah! My game should be really good when it's done and I'm sure I'll have fun making it

I make games for me to enjoy and to learn. I learned so much making Mr. BallGuy! MBG2 was a fail, so this next one will be a lot better. There could be camera movement in it and I'll be using ATL to design the levels.

I hope to learn a lot from this game, and I can't wait to finish it(even though I haven't even got level 1 done yet )

Kind of a long reply but hey, you asked what I was working on and I told you! I've been working on it for 4 days so far!

Don't worry, Eamonn, sharing your ideas probably won't hurt you.
Either:

• You were able to finish the game, and someone else did too. Not very probable, and maybe the other guy will be nice.(As in won't sell his shitty implementation of the idea while giving you the middle finger.)
• You were able to finish the game, and no one else did. The best situation you would get if you don't share the idea.
• You didn't finish the game, and someone else did. At least, there is an implementation of the game.
• No one did the game. You didn't lose anything from sharing the idea.

--- For now, it's just a platformer. The jump code was super fun to tweak, and I'm proud of these tiles.(I think I'm gonna try to upscale them.)
While working on that, I really improved my personal lib. Maybe I could release it.

I've been working on my project for a similar amount of time as Eamonn. I don't really have a ton right now, but soon I hope to integrate levels and enemies. The enemies are there, but you can't really hurt them...
You can check out my game in my signature. Obviously (I guess) it's not mlib, my math library, but the other one: The Epic Adventures of Ninja Dog.
It all started when I was just doodling around in MS paint, and decided I'd try my hand at 8-bit ninjas. After I got around half the body done, I was like: 'Wow! It looks like a dog!'. The idea then evolved from that.
To Eamonn:

There could be camera movement in it and I'll be using ATL to design the levels.

I had a plethora of difficulties using ATL and Tiled. I think it was just because at the time my picture was too small, but I decided not to use it. If you can figure it out, go ahead! A ton of people use that.

I also really liked your MBG2 game, it wasn't that bad. I got more than 100 coins!

Working on a multiplayer C&C thingy without units for the midsummer jam week. Big shout out to Blarget for the awesome art.

I decided to take a break from my bigger project and focus on something simpler, so I'm making a topdown shooter in the style of Raptor. I even stole their ship sprite for now.

wrote my first functional shader.

[[
extern number fartsLeft;
vec4 effect(vec4 color, Image texture, vec2 texture_coords, vec2 pixel_coords) {
vec4 textureColor = Texel(texture, texture_coords);
if (textureColor.w > 0 && textureColor.x <= fartsLeft) {
return vec4(0,fartsLeft,0,0.8);
} else {
return vec4(0,0,0,0);
}
}
]]

As I hate Tiled's (slow!) workflow, and my game only needs a simple tile editor, I threw together a quick level editor for it:



Took about 30 minutes, and *some* of the code is a little messy, but it's pretty easy to expand upon and works well. Can currently save levels, though I haven't added a level loader yet (soon!).

Someday I have to figure out how to read Ogmo Editor's XML levels in Lua. I know a lot of people here prefer Tiled for its features, but using Ogmo Editor is just *so* much faster.

EDIT: oops, screenshot was cut-off on the right somewhere along the line...oh well!

Porting my game project from Python/libtcod to Lua/LÖVE. It is yet another roguelike but heavily influenced by ascii sector and DoomRL along with tabletop RPG's from years gone by (Traveller/2300AD). Current name is Dark Stars Rogue L(ike/ua/ove), and not much to show yet but a spash screen. I'm doing it as a GPLv3 floss project so I'll have it up on bitbucket as soon as I feel a bit more comfortable with the state of it.

That said, here's the splash:


Wanted to have quasi-newtonian movement in space, but in the Python project it was a nightmare to get the AI to do anything much that was reasonable. So this version will have a stutterwarp concept as primary movement (non-newtonian) along with using action points to make it turn based (as in roguelike turns) rather than real time. Spending an awful lot of time writing relatively low level code though, Lua is a world apart in that regards from Python. At least it seems so from my perspective.

Omnivore wrote:Spending an awful lot of time writing relatively low level code though, Lua is a world apart in that regards from Python. At least it seems so from my perspective.

Well, that's also because both lua and love tend to give you as little as possible, it's more a property of the libraries you chose than it is of the language.

1. Generate a grid:
2. Color the grid with x red rooms (hard), y blue rooms (medium) and z green rooms (easy) such that x+y+z = n rooms in the grid (x, y, z can be changed to control difficulty):
2.a) First color it with x red rooms such that no red room has another same colored neighbor (this can be changed if dungeon difficulty should be higher);
2.b) Then for each red room color one neighbor blue and one neighbor green;
2.c) Color the rest of the rooms with the remaining number of rooms for each color.
3. Choose two nodes that are far apart enough and then find a path between them while mostly avoiding red rooms. If you choose a proper x, since red rooms can't be neighbors to themselves and the pathfinding algorithm doesn't go for diagonals, it should create a not so direct path from one node to the other.
4. For all nodes in the path, add their red[, green, blue] neighbors. This should create the possibility of side paths and overall complexity/difficulty in the dungeon.
5. Join smaller grids into bigger ones according to predefined room sizes. Higher chance of smaller width/height and lower chance of bigger width/height.
6. Generate all possible connections between rooms.
7. Randomly remove connections until a certain number of connections per room is met. If n = total rooms, then set n*x rooms with >=4 connections, n*y with 3, n*z with 2 and n*w with 1, such that x+y+z+w=1. Controlling x, y, z and w lets you control how mazy the dungeon gets. If z or w are considerably higher than x or y then there won't be many "hub" rooms that connect different paths, so it's gonna have lots of different thin paths with dead ends. If x or y are higher then the dungeon will be super connected and therefore easier.
8. Reconnect isolated "islands". Since step 7 is completely random, there's a big chance that rooms or groups of rooms will become unreachable.



8. To fix that:
8.a) Flood fill to figure out how many isolated groups exist;
8.b) Pick a group at random and go through its rooms. For each room check if it neighbors a room belonging to another group, if it does then connect them and end the search;
8.c) Repeat 8.a -> 8.b until there's only one group left (everyone's connected).
And that's it, I guess. There's more stuff like adding special rooms, but that's specific to each game, so whatever. But basically I have a few parameters I can control when creating a new dungeon: dungeon width/height; percentage of hard, medium and easy rooms; what color neighbors get added to the original path and the percentage of rooms with >=4, 3, 2 or 1 connections. I think that's enough to have control of how easy/hard the dungeon will be...