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."
Actually, the XBox delay angle has been reported on several places, including very pro-MS/pro-XBox sites.
Boobies never hurt anyone. - Sherry Glaser.
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?
Seriously... This is like the fifth or sixth story from this month's Wired that's been posted to Slashdot. I got it in the mail and read all of these articles weeks ago, and yet they're still slowly rolling in. At this rate, Slashdot will have summarized each Wired article in the current issue individually over the course of the month.
Can't people just go to Wired and read the articles that interest them?
Read: Rabbit Rue - Free serial nove
You're right. Plenty of games have been ported to xbox, or from xbox to PC, and it didn't take years or months to do so.
The game isnt out because noone can run it. It's that simple. It still looks like it's going to require a $400 video card to be playable. I'm not talking super enhanced 2048x1024 with every bell and whistle on, I'm talking to get 30fps at 800x600 you'll need a GeForceFX or R9700.
The market for games that require a 300-400 dollar upgrade just ain't there.
I'm reminded of another FPS from years back (cant think of the name of it, but it was some highly touted Jurassic Park thing) that required a P2, when P2's were brand new and most people still had P200/MMX's. It bombed, because noone could play it, and by the time they had a system to play it on - it was old news.
The same thing would happen if Doom 3 came out today. I wouldnt be able to play it. By the time I buy a new video card, Doom 3 would be old news, and I'd never buy it. Because lets face it, FPS games are in a 'flavor of the week' scene.
Perhaps I'm wrong, and they've gotten the engine to scale down to be playable on average systems. But I'm pretty sure that's a major factor in the wait.
I don't need no instructions to know how to rock!!!!
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.
You're right in saying they didn't evolve.
The development of computer graphics has always been driven by (arguably) intelligent creators.
People like to apply words like evolution to any developmental process presumable for the coolness factor, and in the literal sense they are right (change over time). But it's just silly to imply that CG has evolved in a darwinian sense.
It makes a mockery of the thousands of hours that designers, programmers and engineers have put into developing such systems.
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
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.
It seems to me that some developments aren't obvious until a basic technology is in place. Once upon a time, all telephone calls were manually connected by operators. Most calls weren't automatically switched until the mid sixties. These systems did not spring full blown out of some engineer's forehead. I'll bet some people were thinking about automatic switching in say the thirties but other technologies (which themselves were "changing over time"...I'll avoid the dreaded `E` word") had to get there first to make it a reality. Lots of people worked on it at different times tweaking and prodding and refining until it was mature. The way it works now doesn't even remotely resemble the way it did in the Sixties so it definitely doesn't have one inventor.
Even the most talented engine coders aren't going to be able to tell us exactly how computer generated 3D is going to work 10 years from now. It changes over time and so does the science on which engineering is based. Also, when fundamentally new technologies are going from the whiteboard to prototypes on a bench lots of ideas are tried and thrown out, tried and thrown out, ad nauseum until something sticks. Imagine that!, competing technological ideas going head to head in a fitness race. Sometimes, it's even automated.
But no, technologies are born fully refined and completely debugged from the disembodied head of Thomas Edison which he preserved in his "last" invention.
I won't say the `E`-word though. That might be carrying an argument that already tiresome in the life sciences into engineering.
It seems like a number of games these days look really purty, even though the number of games that actually use the latest hardware or API seems woefully small. It's certainly true that Carmack's one of the key people pushing the industry forward, and that's an important point.
On the other hand, even the prettiest games sucks donkey balls if the AI sucks, or the physics are clunky. I like the suggestion made by another poster--why not code a real deformable physics engine, or come up with a decent AI package for enemies?
On a tangential note, I would be most eager to find out some add-on company bought some balls, some software engineers, some patents and/or R&D, and some cheap, cool X86 or RISC processors and said, OK, we're building an AI/physics daughtercard, and the industry tools to make it work. Oh, and that next-gen cards would be hybrid AI/physics/GPU systems. With PCI Express, we might just have the bandwidth to make it work.
Um, I don't think so. The first first-person perspective game I remember is BattleZone, published in 1983. The first first-person shooter I recall is Xybots (or maybe you'd call it 3rd person), published in 1987.
First, Battlezone is from 1980. There were 1st person games in the 1970s, specifically a few games for the PLATO system.