FS: Do you consider developing a racing game?
John: We used to discuss this issue after launching Wolfenstein. At that time, I was studying some three-dimensional pixel -landfield technology, we have made some simple demonstrations. But we have never taken any measures to put this idea into implementation. But if we really develop a racing game, it is affirmative to a very interesting racing game - such as F-Zero game, not a very serious racing simulation game. Here I have to point out again that there are many effective way of development, but I have my own personal hobby. My hobbies happens to let us develop successful games, but also allow the game to have real entertainment I think.
ID company's next game fs: Do you think your personal hobbies are not driving, but a rocket launcher?
John: The purpose of the game is entertainment. The simulation game has never given me a pleasant feeling, including flight simulation, racing simulation, and all similar simulated games. I know they have certain entertainment in some respects. It is definitely not as I said: "These games are very boring, everyone will like them." However, they are not the type of game I like.
When I just like RGP game, my life became more and busy because of other matters, so I didn't have a lot of time to play games. I have always like simple reel games. Despite the Gauntlet to Quake 3, they have changed a lot, their basic concepts still have no change - you need to run around, defeat the enemy, pick up the weapons, and enter the next level in time. This is a very effective, core, primary game element, I don't think there is anything about the game that makes relatively similar games (such as Quake, Quake 2 and Quake 3) have different styles. . Just like there is always someone who plays racing games and flight games, now there will always have a variety of first person to shoot games. I think each of our projects has got important gains. Our next game may adopt a completely different way, because it is not the same as everyone's ideas. So the next time we may change our strategy slightly, but we have not made the final decision.
Fs: Is it a first person to say?
John: It is likely to be a first person to say games, but it is not necessarily a moving game. Graeme and I have been discussing a lot of game design ideas, and exploring the various options we have now, but we are not ready to discuss any details. It is probably a first person to say games - this is almost certain.
FS: I just thought that if the ID enters the RTS or other type of game field, it may have a very interesting situation - this is like Blizzard stepping from the RTS to the development of the RPG.
John: I am not a real real-time strategy game player. In fact, I have never really sitting down and playing a SLG game. This is not the type I like. It is completely different from the first person to the game - the latter is the most scene game.
Real-time strategy game is one of the best games. You just direct your army, not your own charges in the game. I think we will continue to develop the first person to call the game. From the first person, the game turned to the third person. The game is just a small step, but I believe that there must be some other companies to develop a third-person game, such as the company's more powerful companies in development command strategy.
Unlike many people in the game industry, I didn't make a movie on the computer. And it is this idea that many companies are in trouble. Some game designers think they are directed, and I think you should squat your position. I am just a game designer, not a movie director. This also causes many companies to turn to the third person to say games. Third people are clearly more suitable for special effects such as displaying character animation. Steed will definitely hope to develop more animated games, but this may not be our next game development focus. We have not finally decided, but I will definitely develop the first person to say games.
About the game industry FS: You think that movies on your computer as "the reasons for many companies are in trouble", because there are many 3D graphics card companies - such as 3DFX and NVIDIA - actually "introduce Hollywood into desktop "As their focus of their publicity. JOHN: This phenomenon has a lot of interesting questions - one is the number of people needed to play the power of today's 3D engine is quickly close to the number of movie set personnel needed by Hollywood. Just like now, Quake 3 does not really reach the highest level of visual effects we can make, but only this can maintain a fast enough development speed.
But if someone learned about this year's technical level and said: "I want to make the most amazing, closest movie game", then he can make far more than people The viewing effect level of the game. Relevant technologies are ready. This is the development trend of the game industry in the past year, especially some special effects, such as complete texture rendering of each level, requires scene designers to join in, designing their scenes in different ways, and use models to build all these effects . Such efforts can now be used in a computer game. This is not the case in the past.
When we develop Doom and Quake, we don't have the ability to invest so much. If you need to design a room, a corridor and another room, what you should adopt is: corridors, modules, modules, ceilings. This is everything you can do, so the work we can have has been seriously restricted.
FS: You always support the development of a smaller team; do you mean that children can develop games with their partners in their own rooms?
John: We have not fully explored the possible game types. Many people are dissatisfied with the game industry in many ways, and in fact, many publishers don't like to issue creative novel games. Chris Hecker from Definition 6 has been promoting physical games in a variety of ways; he has developed some novel games. The game he has developed uses a novel idea. He also made a short game presentation. This game will physically act as its core principle, but it is not like "let us physically join a first person design game". It is a very unique game. It uses a good game design and is closely related to physics. But publishers are not interested in it.
This is really very bad, but at the same time, we can also see some positive signs from other things, such as some people earn big money through some simple games they develop on the Internet, such as throughout the year. The downloaded game has been a few thousand dollars income. I feel very excited about this because we can see it clearly, some people can make money through the development of the game outside the mainstream; this situation is possible.
We have successful in the form of a shared software early in the company. Through the initial Keen trilog, our annual income reached 100,000 to $ 150,000. Through the second trilogy, we have reached $ 500,000 a year with the help of shared software. This is still before we enter the formal commercial sales market. It seems that it can be more excellent than at the time. This is indeed very interesting. However, due to the need to occupy costs and shelf space, publishers are not interested in issuing creative novel games. I am sure that there is certain that some new game types are preparing to have an emergence, but they face some resistance.
ID is not a publisher fs: Are you interested in Halo?
John: I have seen its film; it looks very good. The problem now is that it is impossible to judge a game according to the screenshot of a game, because any of the competent rendering engines can render a particular scene. The key is whether each object is seamlessly connected. Bungie has a lot of outstanding game design talents. The time they engage in the game industry is similar to us. They have been like ids on the Mac platform for many years. But now they have already gone in front and have achieved some important results on the PC platform. They behave very good.
FS: Go back to the topic of the new game species and resistance from the publisher, are you interested in helping to promote new game types, such as providing some funds to issue these games? John: I definitely don't want to be a publisher. The ID is discussed for many years on this issue. For five years ago, all game development companies wanted to be a publisher because there were many small publishers at the time. Later, these small developers gradually be merged into a small number of large publishers. We have discussed this problem for a long time, and we have issued the initial quake sharing software, but after consulting with the publisher, we are likely to lose some profits because the distributor's business management is often quite bad. But the last all people have reached a consensus - especially after we see many publishers' mergers - doing a developer much better than making a publisher. Because the software developer is a clear part of this industry, you only need to deliver CD and receive checks via email. However, as a publisher, you need to deal with a lot of chaotic parts that you don't want to touch at all, such as agents paying and selling your products. I don't want to contact this business at all.
FS: or use your influence to convince some publishers, you know, maybe you should take this opportunity.
John: I won't put too much. I will temporarily publish some opinions on such things, but I will not commit a commitment to something I know if there is not enough time and effort. In fact, this is very important for me to maintain a focus on main work. I am willing to pay a lot of trivial matters. This is very important, because so many things may need to take up my time. You can see this importance from a lot of people who face similar situations: Some people keep looking for them all day, making them unable to carry out any job, because people who want to achieve a purpose will call them Telephone and email - access them, or ask them to propose a business suggestion.
Sometimes I feel guilty because there is no more work. Especially in the upcoming year, there are many different hardware design work to be completed, and everyone does not have a clear development direction. I told myself that I should participate in the development of work, helping everyone reach an agreement, because my status makes me more suitable for playing such roles. Microsoft is very happy to listen to my recommendations in many ways, and many hardware companies are willing to do so. They have some respect for me. But during the development of QUAKE 3, I didn't have time to engage in this work.
Now I can propagate some ideas around, but I know, even now, I can't fully invest this job, because I tried and led the entire industry along what I hope to grow forward is actually a full-time job. I haven't arrogally I think that the world will not turn. So I want to say is that I will try to help, but I will not put this work in whole.
About TRINITYFS: Let's change a topic, an article I read mentions that after Quake 3 is successfully launched, you will take more time research engine.
John: I am currently working in a number of engine research.
FS: TRINITY is one of them? I think people have been confused with "Trinity".
John: I have never known why everyone is confused about Trinity. After launching Quake, I started researching new rendering techniques, and people are called "next engine" or other names. Michael Abrash suggests that we use Intel: Name your next project with a nearby river. Dallas has a river name TRINITY, so we will take the name of the next project as "Trinity Engine". Of course, we eventually did not use it for a project, because Quake II is the renewal of Quake I, and I didn't use [TRINITY] on the Quake II, so there is no project adopted this technology. Just now, we talked about the work I did at the time. When starting to develop QUAKE III, I have written six different research engines for different rendering technologies. But this is still just the research work I am engaged in. These include 2 three-dimensional pixel items, 1 spatial distortion project, 1 illuminated project, some of which have become the Quake III engine later.
One thing has always been one of the true strength of ID and my technical guidance. You can see some companies to grasp a smart idea; some games you may be familiar with you can develop game development with a creative as the core. And I am not limited to a specific creativity. As I just said, I have six different tasks to be done when I start developing Quake III. I observed all these tasks and then said to yourself: "These tasks are more simple." But our work for the Quake III is actually a more correct way to make a more correct way. I still have a lot of tasks to be done. But no one will post the magic label above, so I need to personally evaluate all these different tasks.
Three-dimensional pixels and curves fs: You just mentioned three-dimensional pixels, and you are carrying out this research. Do you think 3D hardware companies should pay attention to the acceleration of three-dimensional pixels? How do you think about the three-dimensional pixel game that has been launched, such as the Delta Force?
JOHN: 3D pixel development? I have written some three-dimensional pixel engines in the past. In fact, Shadowcaster - The early version of Raven has used three-dimensional pixel ground in earlier versions. But we rewrite the program to remove three-dimensional pixel ground when we rewrite the program. Display objects with three-dimensional pixel technology can bring some practical benefits, as it can provide you with complete texture and outline details in a variety of ways. I wrote the two three-dimensional pixel engines when I started to develop QUAKE III. I thought I could let them run completely through software, but it can only provide a considerable resolution. Depending on the level you can achieve, and in this case, it is not necessary to use this technology in this case.
I analyzed memory access mode; you can use hardware to make a three-dimensional pixel trace tracker, and the hardware resources it use is far less than what we actually used during the various triangular grating. I think it can provide a more attractive visual effect in many cases. But it is really difficult to understand, so I am not willing to recommend such a technique to people. I know that I have made some introductory demonstrations, but the root cause of the PC industry can reach the current hardware level is that we can learn from the SGI examples. We have the actual evidence of "this obviously can play the role, you have done" the actual evidence of how to challenge and eventually exceed their achievements.
If you want to recommend some new technique, someone will say: "You should make your factory to produce three-dimensional pixel chip, you should try this technology and invest a few million dollars to it." I don't want to do this, because We don't have a complete evidence to indicate that we must use this technology when developing engines, and cannot prove that it can bring enough rendering improvements. If someone really made a three-dimensional pixel development tool, you can get its depth value, mix with the existing triangle, which may be a fun intermediate step, but frankly, I don't think I don't think It takes up so much hardware resources. Since everyone is now looking for a unique situation, some people may use this technology as a propaganda, but they are only only because they don't know how to use this technology. I think it's a very potential technology, but I can't affirm that this is the future development direction, because although these technologies apply to the environmental development, you can use them for some excellent applications, but you still don't know if they are Suitable for role development. You may say: "Maybe you have to build this model with a variant matrix, and then when you run Raycast, let the light turn to the deformation grid when you contact the deformation mesh." But I haven't written this software version yet. Only I can provide a way of working with this technology, and it looks better than you can achieve better, I will recommend this technology to others. I haven't time to do this now, because I have some urgent research tasks need to be done.
But I think this technology has some important potential, which may become an interesting direction. I really want to introduce the technique introduced to everyone to be higher level surface rendering, including curves, subdivision surfaces, etc. Everyone is now interested in these technologies. After using two years, I found that they are not as good as promoting; they have many problems. You often find it directly from the outside. For a developer, you must have to go deep into the inside to see the essence of these technologies.
If you have read the relevant research literature, understand the work mode of the Beyele curve, you may think, oh, this technology is very good, this is the direction of advancing after we are behind the triangular unit. You don't realize all "actual" issues, such as invalid, standard, but degraded edges. You cannot be able to partially divide a curve - unless it increases it to the plane on both sides. You can't connect two intersections by adjusting other parts. All of these subtle shortcomings will eventually cause a lot of inconvenience, especially in more difficult circumstances, it is difficult to fully understand the problem of problems. Three-dimensional pixel technology also exists all these problems, this is why I don't want to go to the stage, depictting the future of the future, because many technologies are very good from the surface, but when you go deep into their internal, it often concludes Conclusion.
Views of competition fs: Since you intend to re-develop research, would this have a certain impact on the next game of the ID?
John: This depends on what measures we finally decide next. Both technologies that I am currently developing apply to existing Quake 3 frameworks, so these new technologies may be added to the next product.
At present, we have got a very important opportunity, we can use this firm, stable, scalable, flexible engine to put our development focused on game design, without considering engine development. . . From a management perspective, this is undoubtedly very beneficial. However, if we do this, the technology I am currently developing is likely to be used in new products in three years, four years or even longer. This is a long period of time for the development of technology. Therefore, we need a really powerful incentive to encourage us to add some new technologies to the next product.
FS: Do you conduct research work by playing other games and products to watch competitors?
John: I often play unreal, I may like to play unreal tournament in the future, although I haven't played this game yet. Also, I will play a lot of demo version of the game. I will not conduct a truly thorough investigation, play all the games on the market. Usually I just look at it when others play the most popular new games. And I will not spend a lot of time to play other games. Of course, I played Quake 3 time beyond any other game. FS: Is there a person in the company responsible for this job?
John: Christian and Graeme will play most of the new games. In the past, Brandon was responsible for playing each game. When he is in the ID, all information will be filtered from him. TIM also often plays games for some research purposes; TIM is a considerable person in browsing other games, while Graeme and Christian have the main purpose of playing other games just entertainment.
CARMACK Research FS: Can you simply introduce your research content?
John: A rendering engine technology I am now developing is a better lighting model. Quake 2 adds an emissive lighting in the calculation, which can bring real light reflection effect, but many people don't like this technology because it eliminates some shadows, so that more objects are in the darkness. Although this is more in line with the actual situation - many people like more obvious shadows. In Quake 3, we re-use a direct illumination lamp, which is somewhat because I don't want to perform finite element radiation calculations on the curve. Another technology will allow designers to render textures and divide outlines anywhere, allowing us to manually improve the display effect of the object. Therefore, there are two main research directions. In fact, they can be combined together, or use together. They have brought me some interesting technical problems. These technologies may be displayed in our next product.
My other research content has nothing to do with the game engine. I am using a web camera to conduct some visual research, such as a head tracking device. I can use this device to add some interesting content in the game. For example, when you come back to your head, the role you play in the game will also make the same action. Or like a window interface that supports holographic photos, you only need to move your head, you can change your image. I don't think that the web camera is used for a quick enough speed, and the above objectives are achieved without significant delay, but this is still a quite interesting technology. You can use facial expressions, etc., increase it to a new level.
I think these technologies will exceed my imagination. All short time discontinuities are caused by changes in the input / output devices. When people start using the mouse to perform various operations, things have changed. I think "Computer Vision" has sufficient practicality, enough to generate "mouse type" to the object. After all, some things can only be completed by mice, but cannot rely on keyboards. This is still in the free guess phase.
There are some interesting applications now, because we now use a strong enough CPU computing ability, you can make some quite complicated calculations. Unlike a lot of image technology, there are still many research issues that have not yet been resolved. Many people are studying these puzzles, but we have no "SGI instance" to know how to combine objects together and improve the speed of operation.
The last research direction I am interested is to achieve more common 3D technology, such as working with other rendering tools (such as HTML) in the 3D world, and small network space that is not associated with a particular game. I think we can build an infrastructure framework. We can develop a first-person design game, but the frame has sufficient versatility, so we can use it to complete other tasks.
All existing software is clearly designed for the first person shooting game, but people can use these technologies to perform other interesting tasks. Generalization is always a cost of paying a lot. You can improve its work efficiency through technical dedication, but in a turning point, you may find that the generalization cost is not very high, and the benefits of having a more general platform have more significant. I think 3D games are currently in such a turning point. We are now different from two years ago. At that time, we still need to consider every short frame to improve a little performance. This situation has no longer exists due to the development of hardware acceleration - especially the outline acceleration, and the speed of faster CPU and larger RAM. I think our generic platform speed can reach more than 90% of the dedicated version, while using this platform to handle a large number of other, interesting tasks.
Comprehensive issue FS: You want to build a permanent Quake world, can people talk and communicate among them?
John: Yes, even don't necessarily be related to Quake. I think there is a lot of interesting work in using media communications. You will see people with each other and meet in a server. There are many potential development directions in this regard. I won't call it a virtual reality because VR has become a name that is linked with death and destruction. There has never been a successful VR company, but I think many of the companies failing have not had corresponding technologies at the time. In many cases, creativity and innovative spirit will be limited by technology development.
Doom is a typical example. We have used a lot of smart, talents and creativity to develop DOM, but all of these intelligences do not have any effects before five years, because the technical restrictions at the time make people can't develop games like Doom. Sometimes the opportunity is suitable for new innovative technologies.
I think there will be a lot of innovative technologies, such as the generalization of the 3D interface, in the next few years. Everyone uses a vague, fashionable method to predict the Internet and the web with a 3D interface at some point in the future. However, no one has truly realized the specific details needed for this goal. It is now possible to implement this timing, but may not.
I am very concerned about these research directions, and I will engage in all these research work. In the end, I will study those most promising technologies in the end.
FS: Do you recommend young people to go to school?
John: If you are going to school for the right reason, the school will be a very good place to learn knowledge. University may be an extremely rich environment. I am thinking about my college life now, I feel that I waste a lot of valuable time. I should use their graphics workstation and other equipment learning knowledge.
If you just have a degree in order to find a satisfactory job, then I don't support it. If you go to school, I will fully support the knowledge and learn new knowledge, then I fully support. Some people have much more in college more than I have listed above. As for me, I have always been the kind of student who will only see the product manual. I like this way of learning, but not everyone likes this. I don't think university is a bad place, but I don't support this point: "You must go to college, this is the only way to success." The reason is very simple, it is not the only way to success.
I think at least for some young, developable industries (such as Internet and game design), can be more important than a resume indicating that your work experience is more important than degrees. I never asked others. "Do you have a degree?" I am more concerned about "What have you done?" If you want to take the exam from sitting in the class or write a game program at home, you have the talent between you. Choose, I think the latter is not lost as a reasonable growth path.
FS: Last question - What game machine you are playing now?
John: The game console that I have frequently played at home is Nintendo 64, but we also have PlayStation, Saturn and some other game consoles.
