A cooperation game where one solves puzzles while others defend him ignited my interest as well. Maybe not very very practical but at least we can all work on the parts we personally prefer ;D.

Imposters hacking their way into the enemy base by solving puzzles, buccaneers to defend, and horsemen to launch an attack.

ya but also remember the work load.. lemme explain why i said multiplayer, a complete professional multiplayer game can be created using:

- only TWO character models, ONE if the game works off colors
- block based maps and some 512 textures
- 8 players
- few gamemodes

either weay, the post will go on forever if a POLL isnt made to vote on the favorite idea/genre/w.e

Yes, the workload would be more, but as nobody has a view on how many will participate with which skills it's hard to say whether or not it's too much .
Indeed we need a poll.

Sounds like a great project.

My idea for a group project:
Co-Op, sci-fi, Diablo-Gauntlet type game.

Art Pros:
Alien worlds lend themselves well to the diversity of contributers where art cohesion could become an issue.
Sci-fi would be a good choice, Diablo has done hell/demons and Gaunlet has done fantasy/dragons.

Code challenges:
Co-Op Multiplayer.
Random map/dungeon creation ala Diablo.
Prefix- Suffix based loot tables for endless variety in loot (Diablo, Borderlands, etc.).



Lots of great ideas so far.

The same poblem as always...
Too high plans for a first project. In my opinion it should be finished in less than half a year, for everything after this first step. It has to be simple and to progress fast, otherwize people will stop working on it and it will never be finished. And if the whole project actually end with not just george doing everything, then there still won´t be more than a samll "core team" with maybe 6 people and with some luck some donators, creating some very small parts.
Thus a quite simple puzzle game or maybe also a very linear hack and slay game could work, but everything else is just completely unrealistic. Yes, there are enough able to code complex things, but how many projects did you already finish? And how many of them have been big? I am pretty sure that the number isn´t high. And if we already fail on small projects when working alone, how should we then be able to finish a big game in a big team?

I btw, could probably help with some Lite-C and HLSL code...

I see the word 'multiplayer' get thrown around too often in this thread but I'm not sure if there's a single lag-free multiplayer game in existence so far, made by A7. I'm not following the forums so much lately so I'm not sure if that's true, but do we have a demo of a working multiplayer internet game yet? I've heard of Anet but I'm not sure if anyone has already utilized it successfully. Over the years I've kinda lost faith in GS multiplayer projects so I would only see differently if I can play a game (made with a7) with any of you online without it being choppy (like I can play Team Fortress 2 with 20+ other people and feel like I'm playing on LAN). The only actual game I can remember is Marble Madness 2, which was pretty fun, but really really choppy

So, if this was the UDK forums, I'd be very open for a game like that, but not here.. Hopefully, with jcl's pet project, or someone mastering Anet, this could change..

well i know that the engine can handle a smooth client-side multiplayer game and i currently work with Anet and know alot of smooth running projects in the works... and if your gonna say a 8 player online game with 4 levels, 2 character models and some basic weapons is aiming too high then this entire thing is already pointless, i've seen one person do that, we could use levels already existing and retexture them, we can put ppl to the test and say create levels and we implement the ok ones, am quite sure we can find 2 decent character models and the rest is code, whether Anet or Native, am only familiar with Anet however...

either way, whatever you guys pick is ok i guess but i dont think we should take and opportunity like this and do a simple puzzle game, we've seen in enough, the market has seen enough and unless you have great artwork thats doomed to fail also, create something that can capture the interest of the majority, a stylized shooter, some arcade like game, a survival game where players hold out against waves of mindless zombies, an old styled multiplayer game where you use bows and such...

Irregardless of this project, I'd really love to play any kind of half finished game like that, if it plays smoothly with 8 players on the net. I'm free to test it now, if you have it on your hd.

well you should be hearing of one real soon and my project is playable but away from public viewing atm but i know its possible.. and smoothness is not a problem in a client-side game really, the main problem you may have left is good hit detecion and there are immense tricks to doing that and theres a multiplayer demo from germanonkul using Anet, its a demo so 4 players but we tested that at 4 players and it was very smooth..

I wish you good luck with that. I know Anet is pretty low level so you have the burdain of making all 'smoothing' functions yourself. Like you said there are immense tricks for doing that and you'll need to utilize all of them for a game that needs the precision of an FPS to be playable. Just slapping a Dead reckoning algorithm won't work. I've read so many reviews of AAA game developers where they state that most effort has been made into making the networking work stable and smoothly.

I'd love to try that germanonkul's demo if it's available for download somewhere..