This is long, I warn you. But, if you happen to finish reading it please discuss.
I remember back when I was looking at parts to build my new PC (in late 2006) Microsoft and others were advertising how DirectX 10 was going to revolutionize gaming. The prime example of what could be achieved was Crysis. At the time I thought that if Crysis was going to be one of the first games to use it then the games that came out 2 and 3 years from then would be even better. Yet, here we are in 2009 (almost 2010) and I still haven't personally seen a game to beat Crysis (or really rival it) in terms of graphical and physics-wise capability. Note: I did not say anything about AI; there may be games on the market that out-do Crytek in the use/complexity of AI algorithms.
As you may know, the gaming revolution did not happen. Sure, we could see tech demos of what was possible with it (heck, I saw an nvidia human head demo of what was possible and I think that was before DirectX 10) but there have not been that many games out that have actually used it. It may have been caused by, in part, Vista's poor adoption rate. There are certainly other factors that contributed as well. Regardless, most games I see coming out today don't look any better than ones that came out back in early 2007 (and they surely don't seem to have come up with too many original/more in-depth gameplay ideas; if there are titles that I have missed regarding graphical and physics advancement CORRECT ME).
Now, that looks pretty amazing. It is almost indistinguishable from water in my opinion. That demo may just be a few frames they were able to render that well, but that level of detail is not currently possible for a full game to be done with. Or, is that level of detail in fact possible for an entire game (to run at an acceptable frame rate that is)? This kind of technology is already possible: http://www.neatorama.com/2006/09/01/photorealistic-rend... it just isn't possible to make that run in a game with normal consumer hardware. Do you think DirectX 11 will actually be a significant improvement in gaming (graphically, physics-wise and AI-wise) or will it fail to catch on/catch on too late, making gamers wait until DirectX 12?
Do you think it will deliver in the GPGPU area? I read this article here : http://gizmodo.com/5252545/giz-explains-gpgpu-computing... about GPGPU (note just DirectX 11) and it sounds pretty amazing. The article read as if physics engines could be much better as a result of this. I can only hope that "they" come up with a GPGPU standard and don't have 2 or 3 or 4 or 5 different GPGPU options (OpenCL, CUDA, DirectX11, etc.) like they have now. For the consumer it would be best if everyone just agreed on one tech (perhaps OpenCL) but it seems like it is going to be OpenCL and DirectX 11, I just don't think Microsoft will try to co-operate.
The only thing I want to see now are some new approaches to game design. Most games I have played in the last 5 years or more all seem to play the same more or less.
These are the things that I want to see even more than graphics and physics upgrades (I like those upgrades, especially the technical aspects of them, but remember, the most important thing about games is having a fun and/or challenging experience):
Longer games (don't think this is going to happen any time soon though as games are getting even more complex)
More in-game options.
I want to see games with more attention to detail. This may be an observation I have made up but it seems that more and more games are trying to appeal to a wider audience; take a look at the Wii, Project Natal, Sony's New Eyetoy and Sony's Wii-like device. This new tech is very interesting and has great potential, but there is an ugly side to it. If one looks at the demographics of who plays the Wii, it will become apparent that many more children and women play the system. What is so bad about that you say? Well, nothing if you are looking at it from their point of view. They have games they can truly enjoy and games they don't have to dedicate a lot of time to. Again, that is great that more people are getting to enjoy games. However, the ugly side of it is often viewed by those who were attracted to video games before they got as mainstream as they are now. I am talking about those of us who enjoyed the NES/Genesis to n64/snes era (I am not leaving you out atari guys, I am just using my age group as an example0. We don't particularly mind learning control schemes with lots of buttons of memorizing lots of combos. We don't mind having to play a game for a few weeks before we can really get in there and play it.
I want to see games with tons of options. I want to be able to go anywhere and do most anything in a game (some games have already gotten pretty close to this). I want there to be so many things to do and features of the game that it may actually take some effort to learn. I want more little details in games.
For example: in Batman: Arkham Asylum you can grapple up some walls and you can grapple up on gargoyles. That is a cool feature. However, if you want to grapple on a non-standard object, like a rafter, you cannot. It would be nice to have it where the player could try and grapple on anything he/she wants to. Yes, the grapple may fail, or it might not be the best place to wait in pursuit, but at least it is there (I haven't finished the game so excuse me if I happen to be incorrect; it is a great game by the way). Essentially, my point is is that I want games to be fully interactive, and to have the stuff there for "more dedicated" players if they want to dig deeper into the game. You could have the same game that the casual and the "hardcore" player alike could enjoy, but there would be a steeper learning curve there for a player if they desired it.
Better stories. Graphics, physics, AI, and all related things are advancing but there seems to be one thing that is being left behind: story. Games need to actually have the stories written by authors specifically for them. Take a great action game with a mediocre story and a great action game with a good story, which one is going to be better? The one with the great story! When films were starting out they had to establish themselves as an art form. I feel that video games are at the point at which they are able to do the same.
What do you think the future of gaming holds? With all the talk of motion sensing camera controllers and 3d screens it will be interesting.
For Graphics and DirectX, the main problem is the fact so many still use XP. Every previous release of DX has been supported by the previous OS, which gives more incentive for devs to switch. But as XP was locked out after DX9, and because of changes to the API, an entirly seperate software path would need to be added for DX10, all the while maintaining DX9 support. As such, DX10 never got any serious development (who wants to built two entirely seperate rendering engines?). The issue is, this still holds true for DX11. As long as XP is around, no future DX release will see significant development.
As for Physics and freedom, I've been begging for a Physics API for years now. A universal Physics API would do more for gaming then any DX release ever could. You could (theoretically) make environments destructable without having to lock into a specialized engine or use specilized code to achieve the effects. You could determine, through the API, how effective weapons are (IE: 9mm hits body armor vs 7.65mm hitting body armor), doing away with damage zones which are frankly, unrealistic. You could deal with enemies in a mountain pass by spontaneously starting an avalanche, without having to have that though of and coded before hand by the devs. Finally, you could have weather effects play a significant role in both weapon accuracy, and vehicle movement. A Physics API would give the user total freedom to perform according to the way he WANTS to perform, and not just the ways the devs want you to.
Of course, Physics must drive AI; the AI needs to know to run from avalances, avoid fire, and deal with constantly changing environmental condition (cover destroyed, Vehicle stuck in mud, etc).
Right now, devs are copying eachother to death; every FPS is based on Halo's health system, and essentially play exactly the same. Every RTS feels the same at the end of the day, every MMO follows the footsteps of WoW. Theres very little innovation left. Worse, as a few companies have basically taken over the industry at this point, every game seems to have EA, Activison, or Ubisoft stamped on teh front of it. Frankly, I think we're nearing a video game crash due to the glut of sub-par games that play exactly the same, a la 1983. In fact, its these same exact reasons why many games can't turn a decent profit; as at the end of the day, every game plays the same as the one that came before it, and people like me have largly stopped buying new games, sticking instead with our older (but more fun) games.
^^ True, but it took a good 2 years to catch on, and had backward compatability with ME and 98, something DX10 lacked. Even then, the base DX9 libary was expanded on several occasions, so all the DX9 extensions practically form a new DX release by themselves!