Phoronix: A Battle For Good Open-Source Game Graphics?
An area where open-source games have traditionally lacked is with regard to delivering high-end graphics capabilities that can compete with modern day commercial games. To this day, many open-source games still look like something that would have been pushed out of a commercial game studio years ago due to lacking proper artwork and a game engine that has an OpenGL renderer that can sustain delivering impressive graphical features with modern hardware...
I just want to throw in my 2 cents:
Enough with the fancy graphics; lets see some gameplay effecting physics.
Remember how in the Crysis demo you can take a jeep, drive it into a building, the building collapes, then you back up into a tree and it falls down too? I want to see more of that on a larger scale.
We already have a commercial quality open source physics engine and with graphic hardware far exceeding the requirements of most open source titles, performance shouldn't be a problem.
I mean, look at warmonger if you need inspiration.
As a developer for the GNU GPL game Warzone 2100 (which was originally developed and released by Pumpkin Studios in 1999 before they were disbanded; released under the GNU GPL in 2004) the problem is not so much updating the engine but rather finding artists to make content for it.
You'd be surprised how much can be done with pure brute force -- high resolution textures, high poly models and some blending. Take the new terrain renderer in Warzone 2100 for example:
The hardest part was getting the textures themselves (namely under a compatible license). Finding people to remake the in-game models is even more difficult.
Sadly things such as support for shaders is not a valid metric to compare game engines by. Adding a pair of fragment/vertex shaders to emulate the fixed-function pipeline is not difficult and has the bonus that it allows you to claim support for shaders. Hardly a useful application, however. Doing something productive and inventive with a shader is much more difficult. Integrating it into an old game even more so.
There's either two things they can do, make Myst/Broken Sword/Fallout-like games, or make Crysis/HL2-like games. Both of these options would make for amazing games, graphics or not, but they also need a real (big) team of people working hard.
Here's what I don't get, though. There are LOADS of 'mod teams' out there, many of them making truly amazing addons for commercial games that could sell as standalone games in their own right. The community is there - it just doesn't know about open source games, or isn't interested.
How come there's no shortage of people for open source projects like IP telephone servers, enterprise middleware, business intelligence, enterprise resource planning, customer relationship management (because when I see those categories I think boring, boring, boring!), but so few to help out with the games?
I don't really think the engines are the problem, the problem seems that the engine makes all these nice particle effects, but the models and artwork are the problem. You can't have a nice looking game without nice models. Nexuiz seems to have mostly solved that except for the player models. Saurbraten, Alien Arena, etc. really need to do a lot of work here. OpenArena is pretty much a q3 clone, which is it's goal, so it's not really possible to complain.