Ryan Feltrin 08-09
August 9, 1999 on http://powergamerz.com/pc/interviews/ridah.htm
A = Ryan Feltrin (A.I. and effects programming)
Q = Power Gamerz
Q: As a programmer im sure you are always trying to think up new ideas to keep the gamer interested. Have you ever thought about persistent online worlds such as a living Kingpin world, which is only filled with other players? Add some RPG elements and it could make for some interesting play. What do you think?
A: I was just thinking about this a few days ago, actually. I think if executed correctly, it could be very cool. The problem with the "persistent online worlds" is they require a large amount of development resources, and therefore a big budget, in order to make it successful. I think we will see a big expansion of this type of game in the near future.
Q: Is there any substance to the rumor that Xatrix's next project may be a sequel to a certain game done by id Software, which was created sometime before Doom and after Commander Keen?
A: I don't comment on rumors, whether they are true or false.
Q: Keeping in mind the recent technology advancements such as curved surfaces, what do you think will be the next significant breakthrough to create more realistic visuals?
A: We are currently limited in the number of polygons we can display at once. I think the curved surfaces is a good step in the transition from large textured polygons, to large numbers of small polygons. We are probably 2 or 3 generations of 3d hardware away from being able to throw such large amounts of polygons at a scene, that it will be difficult to tell in-game graphics to pre-rendered cut-scenes. Hardware transform and lighting is a good step in that direction, the ability to store static geometry in onboard/onchip memory will be the next "big" step.
Q: It seems today that most game companies put 90% of their resources into the visual aspect of games, and not enough toward artificial intelligence. Why do you think this is? Could it be that programmers today just dont have the resources and time to create detailed AI?
A: A lot of this has to do with publisher confidence. If you can show your publisher a good-looking game, they are more likely to push your game harder in it's marketting campaign, give you more time to work on the game, etc. These days, you really need a programmer solely devoted to the artificial intelligence. Also another large factor in the performance of the AI, is to have all the levels completed well before the ship-date. That way you can spend the time making sure the environments within the game don't pose any problems to the existing AI routines.
Q: Do you have any idea whether Kingpin will be released for any home consoles such as Sega's Dreamcast (considering its immediate online capability)?
A: I have no idea.
Q: Finally, what do you think the future of engines and rendering may hold: BSP, Portal, NURBS, other
A: I think we'll continue to see engines that use a mixture of approaches, until the 3D hardware is able to take this task away from the software. Then we'll probably see Portals taking over, or perhaps another hybrid approach. This brings us back to the idea of storing static geometry on-board. If you are able to store the entire world on the 3d hardware, then there is no need to sort the polygons in software, if the hardware is advanced enough to take on this task itself. We may see a transition phase however, where some early cards with this technology will require the software to tell it which geometry to draw, but eventually we'll be able to leave this to the hardware. Having said this, it's easy to forget how far we've come in the last 4-5 years. To try to imagine where we'll be in the next 5 years, with the current level of competition in the 3d hardware and gaming industry, is a very difficult task indeed.
Search:
