Slashdot Mirror
Carmack On Doom III And The Evolution Of Graphics
Toasty16 writes "David Kushner over at Wired has a write-up on the progress of Doom III, hinting at a possible fall release, that is unless Microsoft convinces id to sit on the game until an Xbox version is completed. He also talks to Carmack about the evolution of game engines and the possibility of a "next-generation rendering engine [that] will be a stable, mature technology that lasts in more or less its basic form for a long time." Will this lead to a shift from coders to "technical directors," as Carmack believes? This ties into the Slashdot story awhile back about new titles for sysadmins."
One of the most interesting parts of the article was Carmack's speculations about graphics engines. He sees the graphics engines getting to the point where new ones are no longer needed. After dynamic lighting, how much is there left to do besides minor refinements and optimizations? Carmack remarks that graphics engines will eventually only be done by hardcore enthusiasts. Anyone think that he's right?
Even as computer graphics rapidly approach the quality of those we see on the big screen, CG movies are still a long ways from convincing me they are real. Turing said that a good way to test the quality of artificial intelligence would be to see if it could fool a human into thinking it was a real person. The same concept can be applied to computer generated graphics. We haven't really reached the finish line until CG can effectively fool us into thinking we are looking at a photograph.
As CG in games progresses, software and hardware will need to be increasingly effient (i.e. fast). This almost requires that game engines be written in fairly low level programming languages, ruling out heavy OO design and especially Component Oriented Design (which is the strongest candidate for long-life software).
With each passing year and each passing game, we will be trying harder to achieve the true feel of reality. If engines were component oriented in design, changing one feature such as lighting would not necessarily effect other parts of the engine. In this way it might be possible for a game engine to last more than a few years. However, the fact remains that this is too slow and is impractical for our uses.
Will we ever reach that finish line, fooling ourselves completely? Probably, but certainly not anytime soon.
First of all, let's understand one thing: id Software does NOT need the cash. The company has a history of hits -- they create the technology next-generation games are inspired by, if not based directly on. They're one of the few companies that can spurn the Microsoft money machine and not regret it, because they've been more successful marching to their own tune than just following the easy money.
Second, Carmack has said he's getting tired of making games. But he's not looking to call it quits and retire: he's looking at ROCKETRY, for goodness sake! So here we have John Carmack, one of the most technically saavy minds of our time -- he's a geek's geek, he posts on Slashdot, he doesn't give two shytes about the fame that people would love to heap upon him. Why, then, should the gaming public begrudge him the seed money that could very well open up a new door in rocketry?
Sure, it'll push back Doom 3's release date -- we're still waiting for Duke Nukem Forever, aren't we? Give id Software its due -- let them have the cash, let Carmack make the millions he richly deserves. Because I want to see what Carmack can do when he really applies himself full-time to a REAL-WORLD endeavor.
Yes, the X-Box will have another instant hit if Doom 3 comes out. Is that what some people are hung up on -- MS pulling a Bungie and buying their way to success? Not that it's worked so far -- they have a handful of AAA titles (Halo being the only one I've ever played), and the PS2 still outpaces it in sales.