The real effect of the event being pulled together at the eleventh hour was to damage QuakeCon's status as something more than a big LAN party and game tourney. That's kind of a shame, because QuakeCon has long been distinct from the other big LAN events around the country in its ability to attract media, sponsors, and other folks in the industry. It was as close as we came to a trade show type thing specifically for PC gamers and enthusiastsespecially us old-school types who cut our teeth on FPS games. This year, media turnout was way down, major hardware sponsors like Intel and NVIDIA didn't stage any product launches or big events, and sadly, they didn't even have Kyle Bennett doing a HardOCP hardware workshop. That changed the feel of the weekend quite a bit. I hope the event recovers some of its critical mass next year, because there are an awful lot of big LAN parties, but there's only one QuakeCon.
One new thing we did get to see there was a preview of Enemy Territory: Quake Wars. Activision had a big area set up with a bunch of PCs where people could come and play it. I'm not sure what I was expecting, but the game looks remarkably like a cross between Battlefield 2 and UT2004, with BF2's graphics and environments populated by UT2004-style weapons. You can only see the UT Raptor-esque "helicopter without rotors" thing flying around so many times without having that impression stick in your head. ET:QW may be a little prettier visually than BF2, but it didn't blow me away. Of course, there's much to be said for any game that successfully melds elements of BF2 and UT2004, so I'm not really complaining. I didn't brave the lines to play it, mostly because trying to learn and play a new game in a short period in an environment like that has never worked for me. I'll have to see how it plays when it comes out.
This year, John Carmack's keynote was largely a replay of last year's, with a few new angles on things. Carmack reiterated his delight with the current state of the industry and restated his takes on current console and PC hardware. Overall, though, the content of his remarks were a bit of a downer, because he wasn't bullish on the likelihood of advances in computing capability affecting gaming. Carmack expressed skepticism about a number of things, including the prospects for increased parallelism in CPUs and software, the benefits of advanced audio algorithms, and the possibility of merging GPU-like functionality into future CPUs. He called both interactive physics and wider CPU parallelism "unsolved problems," saying not enough research had been done to show us a clear path forward in these areaswhich makes them distinct from that more successful example of parallelism in computing, the GPU. Graphics chips have long been pulling capabilities into hardware from an established, robust field of academic research in computing.
Even his remarks about graphics echoed the recent refrain from many quarters about computing become a "mature industry," in which advances will come regularly and incrementally, not in spectacular leaps. Perhaps as result, Carmack seems to have moved a little bit in his opinion about that false dichotomy in gaming that perennially fuels discussions web-wide: visual immersion versus gameplay. He said we were getting to the point of diminishing returns from increases in visual fidelity in games, and used the example of computer-animated movies to acknowledge the need for good game design. Pixar's Cars and Chicken Little may be fairly even in terms of their graphics, he said, but one is a better movie than the other.
On a different note, Carmack was upbeat about the possibilities for Armadillo Aerospace, which he called "my own little space program," and he revealed that NVIDIA will be sponsoring Armadillo's efforts in the X-Prize cup. They are working on two different vehicles for the cup, a vertical rocket and lunar lander, which Carmack said they will named "pixel" and "texel."