Gameplay

I'd say that the whole polygon/texture model is inefficient and outdated.

Um, how can you not use polygons for models? Not using 3d models would be taking a step back.

How aboot we shove real people just pulled off from the street into Cds! That will take care of character art jazz and problems! Its engenuis! I don't know why it hasn't been done yet, my idea, My idea! :D

BBG said:

Um, how can you not use polygons for models? Not using 3d models would be taking a step back.

That's just like saying "How can you not use horses for transportation? Not using them would be a step back." Or "How can you not use fire to light your path?"

No way have humans discovered the best way to compute 3D images in using polygons and textures. There are 3D-accelerated computer graphics that use neither polygons, textures, or sprites to produce their output. Milkdrop is one example. The possibilities are pretty much endless when it comes to computer software, and there is definitely some untapped combination of kilobytes that would produce something truly unbelievable on the computer. People have just stopped looking.

BBG said:

Um, how can you not use polygons for models? Not using 3d models would be taking a step back.

You should think of polygons as a way of chopping a model up into pieces that can be quickly rendered on todays graphics hardware. Polygon rasterization (drawing) uses the property that polygons are flat to make the rendering of them much faster. This works nicely when you want to represent flat surfaces, but for curved surfaces (as well really complex surfaces with lots of variations), we sort of use the hack of using many polygons so that theres more of a chance that each of these smaller flat polygons will fall only on an area on the model that is relatively flat (so that the original model is better represented).

There are other methods of chopping a model up. Voxel rendering breaks amodel up into 3D cubes. This has the advantage of being able to break a model up into really small units uniformly in all three dimensions, allowing the representation of not aonly the outside of a model (as with polygons), but the inside as well. Although most of the actual rendering of voxels today is done by breaking the voxels into polygons, this is inefficient in concept, since you are negating the advantage of having uniformly sized elements. A hardware implementation of direct voxel rendering (that is, without using polygons) would be more efficient, perhaps allowing improved resolution in a lot of areas.

Models can be chopped up into patches, which are sort of like polygons, excepty that they don't have to be flat. Modern 3D accelerators from ATI and NVidia can treat polygons as patches by using adaptive tesselation (number of polygons increases automatically as needed) and displacement mapping and vertex shaders. This will give you much smoother surfaces, with much more detail. For example, if you look at the walls from certain angles in Doom3, all the bumps sort of look "flat". Displacement mapping will make those bumps "true bumps", allowing a more realitic effect.

There is also implicit rendering, where you simply have an ideal mathematical model, and you have the ability to determine, for any point in 3D space, if that point lays on, in, or outside the model. In this case, you can achieve the highest quality, since the breaking up of the model exactly coninsides with what you're seeing.

Sorry for the long ramble, but I hope this helps.

Sorry for posting twice, but the first was a long post about something else.

Carmack I think described the FPS shooter a a distinct game type analagous to baseball or tennis-- the basic idea of just running around shooting stuff really shouldn't change much, just as the idea of hitting a ball with a bat shouldn't change much in baseball. In general, I think we are blurring the distinction between a 3D game, and a 3D FPS. That is, a 3D FPS is a type of 3D game. It's very true that there are a lot of innovative things you could add to a 3D game. But not all of them belong in a FPS, just as there are a lot of innovative things you could do with ball games, but putting hoop baskets on 2nd base and awarding runs for hitting a ball into this hoop would be out of place in baseball.

As a concrete example, we might consider adding the capability of summoning your own demons to do battle with the evil demons in the game, a la RPG. It sounds cool on paper, but if this is a core gameplay mechanic, then about 50% of the time, the player isn't going to be fighting monsters, but "dropping" counter demons, and running away. Most likely, the player won't even be around to see all the cool animations and effects, for fear of being hit by a stray fireball in the corridor. So, 50% of the time time, the player isn't seeing the artists painstaking work, meaning that the artists have to do 50% more work. More importantly, 50% of the time, the player isn't really "doing" anything that requires skill-- just a button press, and running while the evil demon is distracted. This means 50% game is effectively filler. The player gets bored.

I'm sure there are ways you could tweak this to get better results, but you get the idea.

You're suggesting that all these FPS'es to come out are basically the same game, and you're right. It's clearly a cheat to many game-players who really have very few options left when it comes to actual new games.

And to say "there really aren't any game styles left to be made" is a stinking pile or the largest kind. The possibilities for game algorithms are endless. However, the developers' bank accounts are not, and they're too scared to veer at all from the status quo.

AndrewB said:

And to say "there really aren't any game styles left to be made" is a stinking pile or the largest kind. The possibilities for game algorithms are endless. However, the developers' bank accounts are not, and they're too scared to veer at all from the status quo.

And with the increased cost of content creation, the situation is just going to get worse (much like we see in the film industry). Still, like you said, it's very unfair say that there aren't any new game ideas-- innovative games ARE being made by indie develpopers, at the expense of production values. It's just a matter of keeping an open mind, and perhaps foregoing some of the polish that you'd get from an AAA title.