I want to do something small and quick. I'm not afraid of art, but I'd rather do a puzzley type game than something that requires a huge amount of artistic investment. I'd love something that utilizes a little bit of procedural asset generation. Anyone have any ideas for a small game that isn't something the market is flooded with already? Or even something you'd just like a better version of if it were available.

I have too many ideas for huge projects I'll never finish right now. I'd rather put those aside and work on something manageable for a bit. So, anyone have a suggestion?

Problem with puzzles is that you can't really make those procedurally, at least if you want variety.

how about "tower offence"? You play the monsters and have to pass the maze Ai builds to stop you!

You mean like Anomaly: Warzone Earth?

Sticking with the theme, Game of Life Tower Defence.

Out of curiosity:

For spring?
Or to sell maybe?
And what does not afraid mean?
Not afraid to make any? Or it doesn't have to be super pretty?
And what is "procedural asset generation"?

(Isn't that your answer right there? Games evolve and grow. With RTS for instance, even if you dream of a game with 40 different units, first make one with 4, and then abandon or expand.
Or am I way off with what you mean with procedural asset generation?)

edit: shot of inspiration

I always thought that stuff like the Doll puzzles on Fatal Frame II were fun.

http://www.youtube.com/watch?v=Vj0Rkeel88A&feature=related

Humans cannot really comprehend Conway's game of life well enough to interact with it in any reasonable time frame. And tower defense sucks...

If you want towers then make Fibre for Android or iOS .

Always thought this was pretty inspiring:
http://www.mojang.com/notch/ld12/breaking/

That was a 24h project, by setting up many shorter goals like that you could accomplish a lot in a whole month.

keep in mind, 24 hours for guys who have done it several times over. Experience makes a world of difference.

Well, regarding 24h (and wow that was tiring), here's what couple of friends and I (3 of us were programmers, 1 was an artist) did in a competition of similar type.
http://dl.dropbox.com/u/29384665/1.avi
Now that's a good example of a team of people who had no previous experience in game development - my friends never made a game (not even a tetris) before, so we all went in fresh.

gamemaker?

gamemaker wasn't allowed (nor did we really intend to use it anyhow)
we wrote it in python (pygame), but we really only relied on bitmap drawing an the IO system it provided

how is pygame?

Well, I don't think it's particularly better than f.e allegro or Java2D. You would probably be better of to avoid it if you ever want to create a commercial game - it's python and that usually means slow, which was actually a problem for us (didn't have much time to optimize stuff).

Besides, after certain basic things are done, such as f.e bitmap drawing, IO system and sound playing, which most SDKs seem to have these days, there isn't any real difference, and it boils down to the programming language.

If we had a choice to do it again, we would probably choose Java2D or XNA, since they're both done in static-typed OO programming languages with good tools. With Eclipse/Visual Studio it's easier to code boilerplate code there than to write slim code in Python, refactoring is also annoying to do in Notepad++.

I guess if you haven't written any code in your life, Python/Lua and other stuff is the way to go, otherwise, please let me have my static-typing and proper OO.

Also of note is that PyGame doesn't do 3D, you can stick an OpenGL binding on top of that but then you're writing raw OpenGL. XNA includes a lot of helper stuff for 3D (many commonly used math functions, for example) and has a certain level of abstraction from the raw draw calls.

I took part in two 72 hour competitions as the artist of the team, I believe we used Allegro on D but that was back when I was in school so a LOOOONG time ago, before all these fancy toolkits came about.

you could help me with my unity racing game; it uses procedural tracks (err... well... it will use them, once i've got the algorithm fleshed out) and is polycount-appropriate for mobile devices. think super offroad, but with a different track every time.

having someone to help with coding would be immensely helpful as it would free me up to work more on art assets; i'm as dumb as a rock but have been able to work things out so far (coding is in javascript and/or c#) so it makes me wonder what a competent coder would be able to accomplish: probably a shit-ton

not crossing my fingers, i haven't been able to lure anyone yet with my sweet candy.

Gamemaker media factory flash weirdness.

You're wanting to do this all in 3D? I don't find unity too intimidating, but I'm specifically thinking something that's not going to take me an age to finish, since I want to have it ready for a resume item by mid summer.

Sounds cool Kaiser, but I think I just have a fetish for the word procedural.